Sprint Guide

belzebub · May 15, 2012, 04:22:20 AM

[a href=\\\"http://www.youtube.com/watch?v=5K_RJedOxZk]http://www.youtube.com/watch?v=5K_RJedOxZk[/url]

The disadvantage of 6-3 stacking is that itâ€™s less intuitive for most people and that it limits the stacking freedom. Especially if you donâ€™t use hold, this gets more complex. To solve this you can rely on skimming on games without (or with low) line clear delay.

If you want to read more about this, I'd like to refer to investigation of caffeine: http://harddrop.com/forums/index.php?showt...2985&hl=6-3

Stacking 2-wide
A strategy on its own is to build up a 2-wide gap on the side, then smash pieces in it very rapidly in a 2-wide way by keeping das preserved. This ensures that a certain part of the sprint has a little lower kpt and gives the opportunity to drop pieces at a way faster rate then normal. Those fast spikes donâ€™t add too much to the global speed and must be performed well to have influence. If you can e.g. put 8 pieces in the gap twice you have 16 pieces that you can fast drop. The amount of pieces is also roughly the percentage of your sprint you affect, here about 16%. Another advantage is that you can easily throw away pieces if you get stuck in stacking. You really donâ€™t have to think about anything because every piece fits in a 2w gap without causing problems. Itâ€™s like having a throwaway hold but then in game, supporting the game. This technique is less efficient if there is line clear delay, as there will be waiting time for every piece dropped.

Stacking freestyle
When you donâ€™t have to look at line clear delay, freestyle is always an option. With freestyle you have the full possibility of keeping your kpt as low as possible, therefore having the least amount of keys needed and the fastest possible time. Iâ€™ve not seen an analysis of a lowest kpt algorithm, but it might turn out that 6-3 will always be very close to the minimum kpt needed to complete a set of 100 pieces. The advantage in this style is that you can put a piece where you want, make holes, can more easy survive missdrops without hold and you have the most freedom degrees to whatever rule you might wanna adhere. In a way all the other stacking strategies above are â€˜specialâ€™ cases of the freestyle category.

http://www.youtube.com/watch?v=Vz1CUa2uZqc

Stacking with Sonic Drop
A couple of weeks ago a new method came on where some application settings in NullpoMino are used (such as sonic drop) and the field was first filled with random piece drops, whereafter the field was then cleared by downstacking through the garbage as fast as possible. There was a lot of controverse about this technique and it wonâ€™t be covered here further.

An example of this style:

stack freestyle to the end of the game
stack so that last tetris gap is formed, then:
- throw away pieces until the I is there
- use the I in hold

A lot of people struggle with the end game. One of the reasons is that the endgame is the part of the game that you play the least as you mostly die in the opening or midgame. When stacking freestyle endgame, avoid ending up too fast in the endgame, your mind will only be able to get it on for 4-8 lines sometimes. Also, if you are not stacking naturally freestyle, you probably are going to loose time if the pieces donâ€™t fit automagically. Some people tend to have great ending skills, look at the lapsilap previous wr sprint ending from frame 94 -> 102:

[fumen]110@eessnbuswbB7UAAEcHYBe/esbMSA6bwNnbxNnbwNbd?QpnbRpnbQp8bvSA5bcDobdDobcDpdkzoblzobkzrbmSADc6?Gob4GkdIwobKw4bDSADcNRnbNRcd1lnb1lCc6SA5bpUoboU?oboUqdZiobYiobYiqbPRA2bcDobdDobcDJdkzoblzobkzOc?9QA2bUKnbVKobUKCdssnbtsobssXcOSA5b5Gnb4Gob4GfdJ?wnbIwobIw3bjRACcNRnbNRXd1lnb1lIc5QA2bEYobEYobEY?obEY1c8eob8eob8eob8eQcsQA3bwNnbxNnbwNGdQpnbRpnb?QpUcqQA2boUoboUobpUCdYiobYiobZiVcaQA0bpUoboUobo?U9cZiobYiobYiccDQA/bNRnbNR2c1lnb1lscNQA1bUKobVK?nbUK7cssobtsnbssfcePAzb4Gob4Gnb5GocIwobIwnbJw0c?JSA5bEYobEYobEYobEYZd8eob8eob8eob8epbPRAeccDobd?DobcDlckzoblzobkzKcMRA4bwNnbxNnbwNJdQpnbRpnbQpQ?cvQA1bcDobdDobcDFdkzoblzobkzTc+QA2b4Gob4Gnb5GCd?IwobIwnbJwXcJQA1bEYobEYobEYobEY0c8eob8eob8eob8e?ScMQA2bwNnbxNnbwN3cQpnbRpnbQpkc9PA0bUKnbVKobUK0?cssnbtsobssncqPA0boUoboUobpUwcYiobYiobZipcDPA9b?NRnbNRrc1lnb1l5cOQA1b5Gnb4Gob4G4cJwnbIwobIwicSP?A0boUmbqU3caimbYi4cMPA0bwNnbxNnbwNqcQpnbRpnbQpz?c/OAxbcDobdDobcDdckzoblzobkz/cDPAHcNRnbNRZc1lnb?1lBd1OAybUKnbWK0cusnbss7cJNAvbEYobEYobEYobEY2b8?eob8eob8eob8eWdqOAyboUoboUobpUecYiobYiobZi9c/OA?7bcDobdDobcDTckzoblzobkz/ctOAybUKobVKnbUKgcssob?tsnbss9cOOAxb5Gnb4Gob4GTcJwnbIwobIwLdjOA8bNRnbN?Rhc1lnb1lEdsNAxbwNnbxNnbwNQcQpnbRpnbQpQdJSA5bEY?obEYobEYobEYZd8eob8eob8eob8epbqPAccoUoboUobpUHc?YiobYiobZiqcePAzb4Gob4Gnb5GpcIwobIwnbJwzcsOAzbw?NnbxNnbwNicQpnbRpnbQp8c/OAxbcDobdDobcDeckzoblzo?bkz+cDPA9bNRnbNRkc1lnb1lAdpOAybEYobEYobEYobEYZc?8eob8eob8eob8ewcNOAxbUKobVKnbUKScssobtsnbssMdCO?A6bqUmboUZcYimbaiQd/NAvbcDobdDobcDKckzoblzobkzU?dkNA5bxNobxNKcRpobRpcdJSA5bEYobEYobEYobEYZd8eob?8eob8eob8epbVPAbcUKnbWKTcusnbsszcjPA+bNRnbNRzc1?lnb1lwceOAxb4Gob4Gnb5GacIwobIwnbJwEdtOAybUKobVK?nbUKbcssobtsnbssCdkOA7bxNobxNbcRpobRpJduNAwb5Gn?b4Gob4GQcJwnbIwobIwPdJSA5bEYobEYobEYobEYZd8eob8?eob8eob8epbiQAncqUmboUecYimbaiecHQA/bdDmbdD7clz?mblzpcjPA+bNRnbNRsc1lnb1l3csPA1bwNnbxNnbwNxcQpn?bRpnbQprcJSA5bEYobEYobEYobEYZd8eob8eob8eob8epbD?RApcNRnbNRkc1lnb1lUcPQA0bcDobdDobcD8ckzoblzobkz?dceQA1b4Gob4Gnb5G3cIwobIwnbJwjcNRA3bUKobVKnbUKK?dssobtsnbssOcaQA0bpUoboUoboU9cZiobYiobYiccfQA0b?cDobdDobcD4ckzoblzobkzhcdPAzbUKnbVKobUKncssnbts?obss1cuPA0b5Gnb4Gob4GtcJwnbIwobIwucMQA2bwNnbxNn?bwN7cQpnbRpnbQpgcJSA5bEYobEYobEYobEYZd8eob8eob8?eob8epbSRAgcoUmbqU3caimbYiMcjQAAcNRnbNR+c1lnb1l?jcERAAcxNobxNNdRpobRpScNQA1bUKobVKnbUK4cssobtsn?bssicjQAAcNRnbNRCd1lnb1lfcJSA5bEYobEYobEYobEYZd?8eob8eob8eob8epbPSAgccDobdDobcD5ckzoblzobkz0bKR?A3boUoboUobpUHdYiobYiobZiPc2RA3b4Gob6GodKwobIwA?cPRA2bcDobdDobcDPdkzoblzobkzIcMRA4bwNnbxNnbwNMd?QpnbRpnbQpNctQA2bUKobVKnbUKBdssobtsnbssYcDRABcN?RnbNRMd1lnb1lUcGQA+b6Gob4G4cIwobKwpc6QA1bpUoboU?oboU/cZiobYiobYiZcJSA5bEYobEYobEYobEYZd8eob8eob?8eob8epbJSAhcEYobEYobEYobEYxc8eob8eob8eob8epbqR?ACcoUoboUobpUGdYiobYiobZiFc2SA5b4Gob6G8dKwobIwq?bsSAPcwNnbxNnbwNSdQpnbRpnbQpwbHTAFcdDmbdD0dlzmb?lzqbNSA5bUKobVKnbUKbdssobtsnbss7bDTAFcNRnbNRvd1?lnb1ltbJSA5bEYobEYobEYobEYWd8eob8eob8eob8esbySA?FcoUmbqUzdaimbYirbDTAPcNRnbNRld1lnb1ltbAAAoeUKh?bNREYsbAAA[/fumen]

A lot of people probably would have ended otherwise and ending up using more pieces. You can rely on luck for the endgame, but it will only limit the amount of games you can finish. You can practice on the endgame by doing freestyle sprints. On the NullpoMino mod you can do a 10L race freestyle to get a grasp on the endgame.

If you really fail every time or more importantly, if you want to take maximum value out of the midgame consistency speed, you can keep building up a tetris gap and finishing with a tetris. The problem with this is that most of the time (when your last tetris gap is ready) you get a piece you donâ€™t need and have to throw it away, so you end up in using one (or more) pieces that you didnâ€™t use. If you use hold for the I, you can then use the I in hold for finishing the game. If you play without using hold you need to quickly drop all the pieces you donâ€™t need towards the I. This may go slow at first, but after a while you get used to it. You can see the I coming in the previews and your brain knows automatically how many simple drops it has to do before placing the I. Whatever strategy you use, donâ€™t try to keep a clean stacking until the last I comes up, just drop them as fast as possible to the next I.

This is in general the case, the endshape doesnâ€™t count, so if in the endgame pieces come you donâ€™t need, throw them away. Also ending the last gap in an unconventional way is not a problem:

[fumen]110@eessnbuswbBLHEcHYCe/erb+SA6b4Gob4Gnb5GsdIw?obIwnbJwpb6SA5bpUoboUoboUldZiobYiobYivbDTA[/fumen]

A good strategy is to see what pieces fit, then throw away everything that doesnâ€™t fulfill. For example if you have a single mino gap, all pieces except O will fit. Another thing to mention is that you sometimes have multiple possibilities to end a game with a piece, then you can choose the fastest way and that can differ from the most conventional way to put the piece:

[fumen]110@geC3jbH3gbA3pbWywdAu48nDFbEcEvoo2AmuMDEFbE?cEoi2TASI4dDQBAAA7eCmyqiAkS0TASYVeDwoo2AiSg/DFb?EmDvg2CEloo2Ay3WeD0oBAAOzqaAkS0TAS4gNE0i2rDFb82?Dzoo2AmSceEF8BAAODd[/fumen]

Now the fastest way to throw away a single piece when you didnâ€™t use hold in your game, is using hold. It will skip the piece and jump to the next piece. It can be an advantage not needing to move it. Letâ€™s say you have a single gap hole and an O coming, using hold then getting to a J will solve your sprint the fastest:

[fumen]110@neD3gbE3pbLZwXAsSceEFbkOEpYQrDFLHSASYxdDN2?aCA7eALhpdA3488AQFVrDkoo2Ahoo2AzOZyDs488AQcTDEF?BAAAzb1lnb1lxbD3FeOxqSAo3MoDFLHSASYJXEPJ98AQ/AA?A7eAOBd[/fumen]

If you had to place the O somewhere else (throw it away), you would have resulted in more key presses, now its hold + rotate + drop. Moving the O to the right would require two taps to get to get the J straight in, or one tap but then the J needs double rotate and a tap or it wonâ€™t fit in. Putting the O on the left needs two taps.

Sprinting efficient (lowering kpt)

Finesse

In order to get the most efficient sprint, you want to get your pieces on place as fast as possible. For every piece placement there are a couple of ways to get the piece in place. You want to minimize your total numer of rotations and moves globally. Therefore you need a good finesse scheme. Especially when you are at high speed levels where you canâ€™t get more keypresses per second than you are at, itâ€™s important to keep the number of inputs as low as possible.

A good finesse scheme to adhere to is the 2-step finesse scheme on the harddrop wiki:

[a href=\\\"http://harddrop.com/wiki/index.php?title=0G_60_Hz_SRS_Movement_Finesse&oldid=9624]http://harddrop.com/wiki/index.php?title=0...&oldid=9624[/url]

Basically it says you donâ€™t need more then two taps and/or two rotations to get your piece in place. It is very important to use both rotations. If you are not using both rotations, start using both rotations as quickly as possible to get used to it. If you use both rotations, consider using 180 rotation, as it will help you to lower your kpt even more.

Very important in the finesse scheme is the use of DAS and the use of both rotations. Important manoeuvres to take a look at are:

Das to wall, tap back
Rotate left/right for SZ pieces instead of rotate + tap once
Das to wall, rotate drop instead of rotate das to wall and drop
I-finesse: you only need max one move and one rotation at all times
Das to wall, tap back, double rotate, this is the most heavy placement as it needs 4 input keys and uses das

Common mistakes are the left and right rotate for S/Z pieces. They use right rotate + tap instead of left rotate for example. Also, for pieces on the left side of the field, you donâ€™t want das tap back, but you can use tapping twice all the time except for the O-piece.

For most people it is difficult to find out what they do wrong in their finesse scheme. To find out you can use the Finesse counter in the NullpoMino mod. It has a counter â€œfaultsâ€. These faults are the number of key presses you did too much. E.g. if you triple tap instead of doing das tap back, it will register one fault. By using this counter it is easy to go through a replay file and see what you do wrong. Also there is a breaking glass sound that plays when you do a wrong move so you can hear when you make faults. We will go further on this in the practice session.

You want to stay away of manoeuvres that take too much time. That is, you donâ€™t want to triple tap for example, or use unefficient ways to place a piece. For some piece placements you might want to deviate from the finesse, just be sure you are aware of the deviation. A reason can be that your succes rate is higher when using an alternative variation.

In the report section of the NullpoMino mod there is a section with the different manoeuvres where you can see what manoeuvres take the most time for you, more on that in the analyse your own sprints section.

Very pro players are so fast that they want to exclude the most complex piece placements. The most complex time consuming placements are those with das tap back and double rotate. This is for example a slowdown factor in the previous Lapsilap WR sprint, where a T is placed on the right side of the stack top down.

Not all the piece placements take the same amount of inputs, so if you can choose for a piece to place it somewhere, you should choose wisely. At one side you want to take care of your stacking, but on the other side you donâ€™t want to waste time. A lot of times you can place a piece like L/J or S/Z without rotating it. If you can have a piece sequence without rotations it can speed up your sprint. When your stack is flat you have the highest possibility of stacking your pieces without rotation. A good example is the opening game that we spoke about in the opening game section. The kpt are kept so low also because no rotations where needed.

Another stat that you want to look at in NullpoMino is the kpt and ekpt stat. It tells you how many keys per tetrimino you needed, with the exclusions of hard drops. A good value in an overall sprint is between 2.7 and 3. The ekpt value is the efficient keys per tetrimino. So imagine you did a triple tap instead of das rotate, the kpt will be 3 and the ekpt will be 2. So the ekpt gives you an idea of what the kpt would have been if you played with perfect finesse exactly the same stacking and sprint as you did. Ideally the ekpt is the same as the real kpt.

A lot of even good players are above 3 and make more then 10 even 20 or 30 finesse faults. If you go at 2 pieces per second and have 3 keys per piece, you press 6 keys per second. If you have 30 finesse faults you pressed 30 keys too much. If your key press speed would be constant that would give 5 seconds. Itâ€™s of course not constant and only for really fast people that are at their key press limit this might get more realistic, but you get the idea.

Also having less keys to press actually makes the total key input sequence a lot shorter and overall it gets less complex, less chances on making faults etc. For example, if you are stacking midgame 4 S and Z blocks after each other, you can just use rotate drop 4 times instead of rotate tap drop 4 times. It makes your burst go much faster and with way less chances on missdrops.

No hold

A very good way to keep the keypresses as low as possible is not to use hold. By using hold, you only waste time. Of course by using hold, you have less freedom degrees while stacking, but once used to stacking without hold, you have a huge advantage. While stacking without hold can be more complex, in the end it gets less complex because you only have to think of the pieces that are coming and donâ€™t have to take into account the piece that is in hold.

When you skip hold, you win key presses as in using a good finesse scheme, but also you lower the complexity of the key press sequence again. You need one key less to think about and the finger exercise becomes a little less complex.

The only reason why you want to use hold (except for throwing away the last piece) is holding the I-piece. You can hold the I piece for making a tetris, either midgame or in the endgame for ending your game. This can be done by holding the I piece of the first bag. When holding the I and you are in the need of an I-piece midgame, e.g. to get a lower stack or to fill a high gap in your stack, try not to use the I directly but try to use the upcoming I in the piece previews, that way you donâ€™t loose time on using the hold button.

Using hold in a sprint can get you into confusing situations. The piece flow changes, the stacking pattern changes, you see more pieces on the screen, you loose focus, etc. Try not to make use of it.

Das preservation

Das preservation is handled in more detail in the chapter about stacking patterns so I wonâ€™t cover it in depth here, just mentioning it as a method to lower the kpt.

When building and stacking, try to make use of das preservation if possible. This excludes an extra keypress and also takes frames away as you donâ€™t need to das to the wall again since the piece spawns directly at the wall. In the opening game you can have some opportunities to use das preservation for 2-4 pieces, but also midgame you can have great opportunities to make use of das preservation.

One interesting way to look at das preservation is how you place a piece. Sometimes a next piece can be shorter to the wall then from spawn position. Then you can use das preservation to position the piece from the wall. This might get more complex for timing. Imagine you put an J on the left and then an O on top of it. The J goes left with das, keep das loaded and then use tap back for the O.

Making use of das preservation is a habit and you should take it into account when playing. After a while it will get a natural feeling to re-use a loaded das. Foreseeing possibilities for das preservation is also possible but hard to do. Once you get used to patterns in das preservation (combinations where you can reuse das) you will spot these and automatically build those blocks without further thinking.

Using your head

Piece previews (focus)

Using the piece previews is key. Use the piece previews on the opening to get a quick opening. Use the piece previews when stacking midgame and use them to get a good endgame.

It is natural to forget about the previews, but try to shift your focus to the piece previews and know where to place the pieces before they spawn. The moment it spawns you should immediately be able to place them. Some people practice this blindly by disabling the active piece view. This goes hand in hand with a good finesse scheme so you can instantly place the piece when it spawns.

Very good players report that by shifting focus to the preview they get much better results. Itâ€™s probably a bit like professional racers, they are always 100m ahead in a turn and race, somehow they think about it then their unconsciousness takes over and acts like itâ€™s supposed to happen.

Putting yourself over the limit

Itâ€™s very important to keep pushing yourself to get better results and not getting slow by not trying to push it harder. However keep it fun, having fun is always important. The key is to push your brain to learn and it needs new inputs and changes to learn new things. So try different things to keep your brain focused.

When you feel you are on a plateau, go do some 10L sprints, do some speed races, try to improve finesse, look at your stacking, try to work on your openings or to get bursts. It will throw everything upside down at first, then try to work to get it back consistent and you will improve in the end. Try to play against other people or play under a bit more pressure to keep you going. You can improve a lot of speed by playing good multiplayer matches that will push you over the limit to stay alive or try to win.

Also when you are playing, you can try to set a goal, like â€œI want to get sub45â€. See what speed you need. Try to work to get there. For example play some speed races or 10L races to get used to the speed. If the speed jump is too high you canâ€™t jump immediately to it, it has to come by playing a lot and then you will gradually go faster. Try to see where you have limitations and try to eliminate them. You probably canâ€™t do perfect finesse instantly, but you can improve gradually. Donâ€™t get frustrated, the learning curve can be high. Also itâ€™s not because you seem to hit the same time all the time, that means you are not improving on some levels. You might get better in finesse, you might get better in double rotation, you might learn new patterns, etc. Also you might just be fast but happen to have games with a lot of rotations/moves.

Also, take a look at the ideal time to see if the ideal time beats your record. If the ideal time beats it, you are capable of doing it by playing more consistent and chance

NoManual · May 15, 2012, 04:23:31 AM

This is amazing! Will take a closer look when im done with exams! Good job belzebub!! 5 stars

belzebub · May 15, 2012, 04:24:56 AM

Eliminate hesitation

Hesitation is really bad and reduces your time a lot. Hesitations can come in different forms, but a lot of them can be reduced or avoided.

For example, when you donâ€™t use hold, you can never get into situations where you used hold and see too much pieces coming and donâ€™t know what to do. By not using hold you can avoid already a subset of hesitations.

Then, sometimes you hesitate where to place a piece as it has multiple landing options. Ideally you should know on beforehand where the piece has to go. If you donâ€™t, donâ€™t go wasting time on finding the best place, just drop it where it matches. In your brain, your brain has a certain view of looking at the stack and where a piece should go. Sometimes itâ€™s not the best place to put it, but at least you didnâ€™t lose time. If the stack gets worse, your brain is learning that it should prefer another place. Most of the time you can just go on and stack further, as for every piece there is always a way to put it. Not having a place to put a piece doesnâ€™t exist, there is always a place.

Of course this has two sides. If you just drop fast without thinking you will always end up in bad stackings. Then of course you should do more thinking and try to improve your thinking time instead of foolishly dropping pieces. But if you are smart enough (think fast enough) and your play speed is slower then your think speed, you can usually go much faster by elimination hesitations.

Another way to reduce hesitations is, if you see block combinations and decide to use them, then stick to them and just follow the combination. Also if you are thinking it wasnâ€™t the best option while doing the combination. Just stick to it and go on.

Practicing sprint

Nullpomino line race aid mods

There are a couple of mods in NullpoMino that really can help you while practicing your sprint. We will give a quick overview here.

border color: in the mod there are settings for HARD LIMIT and SOFT LIMIT. These will color the border and the speed indication. If you drop below you soft limit, the border becomes orange. If you drop below your hard limit, your border turns red. If your border turns red, itâ€™s very hard to recover and get back to orange/green, so if it stays red and you are in a difficult stack situation itâ€™s better to restart. Experiment a bit with the settings. Some people set their hard limit way higher (like 10-20ppm) then their target record, so they can see after the start how long they go on fast speed after the opening game, and their soft limit a bit higher then their target speed (like 5-10) or on their target speed to keep a look at the midgame. When you are then going to your endgame and are still in orange you know you are having a big possibility to break the record or set a good time. This simple mod helped a lot of people with breaking their record already.

Ppm indication: when using ppm (pieces per minute) instead of pps (pieces per second) you speak about different numbers. The difference between 120 and 125 ppm is much more clear and impressive then the difference between 2,0 and 2,08 pps. It also gives a good understanding about the subtilities and you can view progress much faster when using ppm. Remember from the start of the article that the faster your sprints are, the more ppm you must gain to get better sprint results.

Finesse counter: on the right bottom, there is a finesse indication. Before starting the line race, you can choose to display kpt or faults. The faults are the keypress overheads, so the keys that you pressed too much to get a piece in place. For example if you rotate tap an S instead of rotating it to the other side, it will give you 1 fault. Those are not really â€œfaultsâ€ you can chose your own finesse scheme, but they are deviations from the 2-step finesse as displayed on the harddrop wiki. Every fault can be interpreted as a key that wasnâ€™t really needed. It counts 2-step finesse and doesnâ€™t look at 180 rotate (yet, ask Polaris if you want this). On the other hand it has a kpt counter, this counts the keys per tetrimino, except the holds (youâ€™re not supposed to use hold anyway) and the hard drops. A good value to strive for is 2.7 and generally a good sprint has a kpt between 2.7 and 3.0. If higher you probably have a bad finesse scheme or made really bad stacking choices. If you have one of the two values and want to know the other value, open the replay file and search for finesse.

Reporting functionality: when finishing a line race, you can press the F-key to open the generated report in your browser (works best in FireFox for some reason). The report contains speed graphs, fumens of your sprint, analysis of ideal time, etc. More about this in the analysis section.

Speed race
There is a game mode called speed race in the mod, where the goal is to place a certain amount of pieces without dying. You die if you top out or are out of health. You start with 100% health and you lose health if your current speed drops below a speed level. In this mode you can get used to quicker play. In the beginning you will stack badly with a lot of holes, but after a while you get used to the speed and you start stacking better and your brain starts processing faster. When I first created this mode, my sprint record speed was 120ppm. The day after I completed a 60 piece game with 140ppm. I switched back to sprint and almost immediately broke my record and went sub50. This is not a golden egg, but the speed race can help you stimulate your brain, put yourself together and fight all the time. Also it is less hard then sprint as you donâ€™t have to keep to clean stacking and can go on. Best part of it is, itâ€™s a lot of fun to play, well most of the players who tried it found it a lot of fun. Fun to know is that the idea of speed race originated by reading a post of Blink, where he said that after a while you get used to higher speed.

Donâ€™t only play speedrace. Taking care of good stacking and other things is important too. When playing speedrace you can put yourself over a mental limit and pressure yourself. You need to get that speed consistent, with good stacking to get the results of it. Just always know what you are doing. Same for gravity.

Gravity
Another way to force yourself to play faster can be to put the gravity as high as you can manage. This will force you to put your pieces at a certain frequency and as fast as possible. This is a little bit less fine grained then speed race as it is piece per piece and not all pieces take as long to place as others (complex placements vs. simple drop), but it is a good way of pressuring yourself in a quick way. Another disadvantage is that if your stack gets high, you will have much more impact of the gravity then when your stack is low. This can be a real disadvantage for people that tend to stack high first before they start to tetris things away. Same reasoning as for speed race, donâ€™t focus only on finger speed only, stacking and knowing what you do is very important, but this can be a great way to kick yourself once in a while.

Finger training
If you look at the top players, a big part of them has fast fingers. Not only because they play tetris, but also because of other reasons, like, a lot of them play StepMania or play an instrument like piano. StepMania is a finger exercise game where you have to press the right buttons at the right time. Playing StepMania will help your finger speed and flexibility and train your brain in fast and tiny muscle movements. It is not perfect of course, because the combinations donâ€™t map directly on tetris manoeuvres, for example, pressing left and right together is not something you will do in tetris. However it helps for finger speed and hand eye coordination. No wonder a lot of pro players also play StepMania.

Another way found (very) effective by some players to train finger speed is to replay sprints. For example in the mod you can choose to replay LapsiLap previous WR sprint. Learning the sprint by heart can be done really quick (takes only 5-10mins). Replaying it over and over again will train your brain and muscle memory and you will go faster and faster. However most people will have problems the moment that they get at their finger speed. For example, a player with 55s record can get his replaying maybe at 45s. However this is still far away from 20s. However this is not think time, as his think time is almost 0 (he knows by heart where to place the pieces). There is also no time wasted on finesse, because the player replays it with perfect finesse. So whatâ€™s left is pure finger speed that results in the ability of the player to place the pieces where they need to come. Some complex manipulations like das rotate drop can be significantly slower for a player that just starts working on his finesse compared to a pro player. Also really fast players have so much feeling and training in their fingers that they can get good results at very low das settings. Training by replaying known sprints can help a player in getting a grasp on this difficult finesse placements without all the hassle of the stacking.

No hold play
Playing without hold can be a real hassle if you are not used to it. Most people use hold all the time to get a clean stacking. It can be difficult to force yourself not to use hold. A fun way to get used to playing no hold is to play cultris. However Cultris is more about downstacking then upstacking, but itâ€™s fun and you get a grasp on no hold playing. Be hard on yourself and just try to never use hold. Once you get used to playing no hold, it might become confusing when you start using hold again, but only for a few minutes. Playing sprint with no hold will definitely be handy also in your normal games, as you can will be stacking much faster just by not using hold.

Stacking patterns
One way of getting used to patterns with JLOI is to use the Combinations mode line race in the NullpoMino mod. To start this go to start -> line race -> combinations. The combinations mode will only give you pieces that fit into a combination with JOLI. Those are the combinations given in the combination section of this article. You will get used to the combinations and see them in a game. This is an example practice using this combination sprint mode:

[fumen]110@feYimbaiwbj6UAAEcNRnbNRid1lnb1l7b2RA3b4Gob?6GkdKwobIwEcGTAEc6Gob4GudIwobKwtbjSAEcNRnbNRmd1?lnb1l3byRA5boUmbqUndaimbYiDcFRAAcWKnbUKJdssnbus?YcdQA1bUKnbVKobUK4cssnbtsobssicOQA1b5Gnb4Gob4G2?cJwnbIwobIwkcFTAEcWKnbUKzdssnbusqbdSA5bUKnbVKob?UKidssnbtsobss0bOSA5b5Gnb4Gob4GgdJwnbIwobIw2bGR?AAc6Gob4GKdIwobKwVcjQAAcNRnbNRCd1lnb1lfcGRAAc6G?ob4GNdIwobKwSckQA/bxNobxNEdRpobRpcc2PAzb4Gob6GG?dKwobIwmcqOAyboUoboUobpUacYiobYiobZiBdaOAwbpUob?oUoboUScZiobYiobYiLdGPA8b6Gob4GpcIwobKw6ckOA7bx?NobxNgcRpobRpEdyPA1boUmbqUDdaimbYircFPA8bWKnbUK?ocssnbus9ctOAybUKobVKnbUKdcssobtsnbssAdWOAwb4Go?b6GocKwobIwHdlNA5bWKnbUKMcssnbuscdtMAubUKobVKnb?UK5bssobtsnbssod6MAtbpUoboUoboU6bZiobYiobYimdqM?AuboUoboUobpU2bYiobYiobZipdmNA5b6Gob4GNcIwobKwZ?dDNA5bNRnbNRFc1lnb1ljdSMAuboUmbqUGcaimbYivdPOAw?bcDobdDobcDackzoblzobkzDdaNACcpUoboUoboUsbZiobY?iobYifdPOAwbcDobdDobcDackzoblzobkzDd6OAFcpUoboU?oboUPcZiobYiobYi5cGPAac6Gob4GHcIwobKw+ckOA7bxNo?bxNccRpobRpId2NAvb4Gob6GecKwobIwSdjOA8bNRnbNRac?1lnb1lLdGPA8b6Gob4GocIwobKw7c9OAybUKnbVKobUKjcs?snbtsobss6cVOAxbUKnbWKqcusnbssGdiNA5bqUmboUScYi?mbaiYdSNAwboUmbqUYcaimbYibdjNA6bNRnbNRIc1lnb1lf?dCNA4bqUmboUFcYimbaimdnMA4bdDmbdD6blzmblzxdyLAt?boUmbqU6baimbYi8dmMA3b6Gob4G7bIwobKwtdDMA3bNRnb?NRzb1lnb1l3dSLAsboUmbqU0baimbYiDeaNAubpUoboUobo?UIcZiobYiobYiXdjNAOcNRnbNR0b1lnb1lfdmMABc6Gob4G?ubIwobKwwdEMA2bxNobxNvbRpobRp6dWLAqb4Gob6GxbKwo?bIwEeCMA2bqUmboU2bYimbai3dnLA2bdDmbdDrblzmblzCe?+MAub4Gob4Gnb5G4bIwobIwnbJwpdOMAtb5Gnb4Gob4GubJ?wnbIwobIw0dDLA1bNRlb1lNRlb1lReaOAwbpUoboUoboUac?ZiobYiobYiDdvOAPccDobdDobcDFckzoblzobkz5caPAGcp?UoboUoboUYcZiobYiobYivcmOAZc6Gob4G+bIwobKwIdDOA?7bNRnbNRUc1lnb1lSdSNAwboUmbqUVcaimbYiedqNAwboUo?boUobpUIcYiobYiobZiVdaNAubpUoboUoboUAcZiobYiobY?ifdDMA3bNRnbNRtb1lnb1l9dDLA1bNRlb1lNRlb1lRejMA4?bNRnbNR4b1lnb1lxdjLA2bNRnbNRmb1lnb1lFejKA0bMRgb?0lob0lZeyOAzboUmbpU8ezcaimbYi9cvPAzbcDobxNobwN1?ckzoblzobkzlcGPAQc4GNRobMRUcIwobKw7cjOA8bNRnbNR?gc1lnb1lFdOPAzb5Gnb4Gob4GscJwnbIwobIwwcVOA7bUKn?bWKfcusnbssHddOAxbUKnbVKobUKacssnbtsobssEdiNA5b?qUmboUQcYimbaiadDNA5bNRnbNRDc1lnb1lldWMAsb4Gob6?GFcKwobIwudmLA1b6Gob4GobIwobKwCeNNAvbUKobVKnbUK?HcssobtsnbssZddMAtbUKnbVKobUK2bssnbtsobsssdKLAr?boUoboUmbYigbpUlbYiobZiHeCLA0bqUmboUgbYimbaiPej?KA7eAGKAsbIwkb5GMRgbKwkbMRbeGJAwb6Gob4GdeAAAsb4?GmbNR6GjboUNRoUlbrU4GkbpUNR6GgbUKgbPR6GVKgbRR4G?VKgbRRoUVKgbPRqUVKgbpUNRYKgbsUUKNRUKgbpU4GNR7Gg?bNR6GqU4GgbNR6GoUeDgb4GVK4GNR4GxNgbWK9Ggb6GqU5G?UKgbNR4GoUNR4GVKgbNRoU4GNR4GVKgbqU6GWKqbAAA[/fumen]

In order to practice with other pieces like S/Z/T you can use practice mode. In practice mode, you can enable and disable pieces that you want to practice with. Practicing with S/Z/T can be really handy in getting your finesse right and stacking with those pieces.

The sprint manifesto

Iâ€™ll try to put some basic rules for sprint in a manifesto:

- don't (ab)use hold
- use both rotations
- use das
- use your best finesse
- play consistent
- use the piece previews, know whatâ€™s coming
- play a lot, even more
- have fun
- know what you are doing

Analyzing your sprints

Iâ€™m gonna keep this one short. If you are serious about analysing your sprints, you can take an example of the Microblizz sprint analysis. It includes good examples of ways to analyse your sprint.

Another thing i want to mention here is, try to find people of roughly the same sprint time as you are. Then compare the reports. Maybe you have better finesse or you are faster in manoeuvres but you used too much hold. Finding differences between reports is actually a lot of fun. You might figure out that your das movements are rather slow compared to somebody with roughly the same sprint time and can work on that. You might find that you are much more inconsistent or using way too much pieces or are sprinting faster but less efficient (kpt/ekpt), etc.

Using the reporting tool

With the latest mod, you can just press the F-key to open the sprint report after your sprint. This will open the sprint report in your browser. The report contains tabs with in each tab a lot of information about your sprint.

General statistics

There are some general statistics that are of obvious interest:
- the sprint time
- the pieces per minute you played (the speed)
- hold used: ideally zero
- pieces used: ideally 101-102
- finesse counter: ideally 0

Speed
There is no good time with a good sprint speed. The speed graphs are in the speed tab. The first tab gives the global and local speed during every game frame. The local speed is calculated on the last 10 pieces you placed and can vary. The global speed is the total speed until that point and the speed that will count when you finish. Drops in the global speed can be explained by looking at the local speed. The best sprints have a good consistent local speed with a few high spikes in. You want to avoid lower parts in the local speed as they count hard for the global speed.

On the x-axis are the frames. You can view the frame number in NullpoMino mod under the piece previews. If you enable frame step in the options, you can go frame by frame through the replay. Another and quicker way is to go to the replay tab and just type in the frame number in the fumen.

Under that graph are two graphs with the same data. Those give, for each piece, the time used to place the piece. So you want this to be as fast as possible. The first displays the amount of frames needed, so you want the amount of frames to be as low as possible. The next one gives it in milliseconds. You can investigate a piece that was slow by just clicking on the point in the graph. It will give you the number of the piece. You can then drag the dialog with the game fumen in next to the graph and just go to that piece by typing the number in the fumen entry field. You will quickly see the piece that was too slow. This can e.g. be a piece that needed das tap back and double rotation.

Time vs ideal time
On the statistics tab, you have interesting information about the speed too. You can see the average amount of ms/frames used per piece. This number should go gradually down when you get faster. It gives also good information on the pieces that were slower then 95% of the placements in the sprint in the category â€œtoo slow piecesâ€.

The sprint is taken and all the placements are taken into account, then a 95% certainty interval is taken. This means that 95% of all the placements in the sprint are between this timings. Only 5% is slower or faster. If those 5% of pieces were all faster, you donâ€™t have too slow pieces. However most of the time, you have slower pieces and they are displayed.

Interesting is the ideal time that is calculated. This is the time that you would end up when you placed all the pieces within the 95% interval, so without the too slow pieces and without really fast spikes. Usually the ideal time is a bit better then the real time. If you have really slow pieces and outliers, you need to more consistent and you will end up more in the approximation of the ideal time. If the ideal time is under your sprint record, most of the time, you are going to break your record soon.

Remove manipulations that you don't need

On the manipulations tab, there is a representation of the manipulations you did. You can see what manipulations were frequent and how much time they took. You want to avoid slow manipulations like tapping three times. If you see a lot of triple taps, thatâ€™s bad. How bad? Well look at the graph on the bottom of the page. It displays how much time each manipulation category took. Compare your slowest manipulations with your simple drop/tap/rotate. If they are insane larger you know where you loose time. Get rid of complex and long manipulations if you donâ€™t need them and take a look at finesse. Some manipulations can take much more time then needed, for example when you are not used doing them. A good example is das rotate drop. If you learn to do this late, it will be slower then the rest of the manipulations. You can try to practice it a bit then if you use it frequently, it will help overall. Also for example a simple das tap back can be new if you never used it. Try to practice it for a while and it will go much faster. For example just go to practice and place every piece with das tap back as fast as you can. It will come naturally if just use it always when playing anyway of course, but itâ€™s a good thing that you are aware of your finesse and helping it a hand can do no harm.

Keeping your keys per ekpt as low as possible

There is another example at the manipulations tab: the frames per key per piece. This is the amount of frames that you needed for every keypressâ€¦ per piece. The idea is that a simple drop of a piece should be faster then a complex movement like das tap back double rotate. If a simple drop takes the same amount of time as this complex movement, you really lost time at that drop. Therefore the total time needed to place the piece is divided by the number of keypresses. By doing this, you get a good statistic for the efficiency you placed the piece with. If this number is high, that was inefficient. Again you can click on the high nodes and see what piece was placed inefficient. You can then use the dialog of the fumen to just go to that piece and look what went wrong there. Probably there was some hesitation going on because the next piece would not fit naturally or so.

Example: sprint analysis of Microblizz WR

A good full analysis of a sprint can be found here.

http://harddrop.com/forums/index.php?showt...amp;mode=linear

Conclusion

Sprinting is a craftsmanship on its own. At all times try to have fun. Read and think, then do your own stuff. Pressure yourself but also know what you are doing. Remember that you will have to play a lot. Way more then a lot in fact. Donâ€™t get greedy or frustrated, you will get faster in time.

--
Last: Feel free to send me additions/remarks/corrections to this guide and I'll include them.
Very last: if you break your record with using the information here, feel free to add a reply on this thread

StS · May 15, 2012, 05:44:39 AM

Belzebub, you beast. Your guide is longer than my project report, which I took an entire semester to write. Goddamnit. =D

This is looking like proper quality work. Thank you for sharing!

vipjun · May 15, 2012, 08:45:12 AM

this is nice, tag to read for later

Shuey · May 15, 2012, 09:08:30 AM

belzebub, this guide is absolutely fantastic! You deserve mass respect for not only sharing something like this with the community, but for the amazing amount of detail. Nice job!

Paul676 · May 15, 2012, 09:28:34 AM

Just read it all - firstly, I'm putting it in the most useful threads list, right near the top so people can see it.

Secondly - excellent guide, extremely thorough - I can't think of a single thing you missed!

I like your manifesto's "don't (ab)use hold" - maybe noting another view in the hold section, just saying something like "some people like using hold because it helps cleaner stacking, more predictable patterns, and the ability to use regular patterns more by using hold intelligently to change the piece order. This does not mean that abusing hold (using it too much) is good, however." would be a good idea.

xenoslash · May 15, 2012, 09:34:59 AM

"Piece previews (focus)"
This!

Anonymous · May 15, 2012, 10:08:11 AM

Pro.

caffeine · May 15, 2012, 11:42:12 AM

Another excellent quality and very thorough post! I can tell you put in a lot of time into this, and I think a lot of players will benefit from it, so thanks!

Paradox · May 15, 2012, 11:55:52 AM

You can quote this in a reviews section on your OP lol:

This is the best and most accurate 40 Lines guide to date. It is very fact and experience based. Great work belzebub!

also don't forget credits

I spotted a genius opening build in the guide somewhere.

belzebub · May 16, 2012, 07:57:21 AM

Aww thanks guys! Yeah, I put a lot of time in this. I think it can help a lot of people and it answers a lot of questions that keep coming back.

It's a bit a pity that it didn't fit in one post :-D

@Paradox: You showed me that opening and at that time I recognized it every time I had it in my own game, so lot of thanks for that. I saw it with other players too and was not sure if you "invented" it, so at the moment of writing, I decided to give you credit by naming the image on the server =))

Alexsweden · May 16, 2012, 08:33:20 AM

:O

At least in my heart you will be PotM next this comming month for this!

SirJeivus · May 16, 2012, 11:34:21 AM

belzebub, this guide is extremely well-written, detailed and engaging but at the same time it is focused and to the point. Very commendable work!

Hobs · May 16, 2012, 07:57:35 PM

Beautiful. Thank you for all your hard work and dedication.