23DenaliBDE
Well-known member
I would hate watching someone slow play to the end of the shot clock every ball in order to run out time so they knew they were on the last rack and could not lose. Awful
which favors the higher-rated player, winner break or alt. break?
- Thread starter evergruven
- Start date
tomatoshooter
Well-known member
Yeah, you'd have to find a way to limit that. What about the English pool, where the tiebreaker is who can clear a rack of 6 balls in the least amount of time? They finish in like 25 seconds or less.
There has been a recent discussion of using a chess clock. With a chess clock, the other player cannot run out your time -- he can only use up the time on his own clock. I think a common time is a really, really bad idea at pool.
They sort of do a common clock at snooker with the shoot out. Each match is a single frame and the frame is limited to 10 minutes total time. However, there is a shot clock that is 15 or 10 seconds. At snooker. The players tend to move quickly. There is a little running out the clock but with such a short shot clock it's hard.
Last edited:
tomatoshooter
Well-known member
I've seen that suggestion and I like it. A player can bank a little time to ponder more difficult shots. With a standard shot clock there is no incentive to play quickly.
That's the thing with stats. Depending on how you want to view them, they can either support or discount your theory.
Case in point, the way Mike chooses to present the stats clearly show that 'overall', the Toyota group won no more racks in either format. However what it doesn't disclose is how many (if any) matches were more closely contented. Maybe what happened in the alt break tourney was a portion of group Toyota breaking bad (subsequently losing badly) and another group getting to the hill against Ferraris....?
Here's what I know. If we throw out the statistician logic that believes you stand nearly the identical chance to win the lottery if you don't play it. Then we can grasp the reality that alt break gives the opponent an opportunity to win racks they wouldn't in winner break. It really doesn't need to go beyond that. Sure, if you stretch the sample size long enough, all stats will smooth themselves out.
In a set by set analysis, winner break clearly favours the stronger player. Just like a hand rack favours the stronger breaker, or tighter pockets favour the stronger potter.
Now on what is more entertaining...? I personally rather see players actually play one another. I like seeing players under pressure. Sure seeing someone string 7 racks together would be cool. However it's an exhibition at that point and not a competition. Doesn't matter if Gorst is playing SVB or myself. I'm actually more inclined to say that open events should be winner break until the quarters or round of 16. Burn off all the dead money and then see the strong trade blows.
If the individual shot clock idea were put in place, players like Corey would find a way to abuse it.
Its easier for pros to force defensive play, but against dead money, its just sloppy safe returns.
I am not saying its a bad idea, its just one where the forethought hasn't been fully developed for potential abuse.
I hope to see the day when pool engineers find a way to make players have 9 golden breaks in a row.
You can argue time as a future variation. I will blame the poor development of equipment on the product engineers.
Even high school physics students know reactions should be predictable with the same initial conditions.
Why cant pool engineers solve the golden break? The Harriman Research Tech team is working on it but without support or endorsements development is slower than it should be.
Its easier for pros to force defensive play, but against dead money, its just sloppy safe returns.
I am not saying its a bad idea, its just one where the forethought hasn't been fully developed for potential abuse.
I hope to see the day when pool engineers find a way to make players have 9 golden breaks in a row.
You can argue time as a future variation. I will blame the poor development of equipment on the product engineers.
Even high school physics students know reactions should be predictable with the same initial conditions.
Why cant pool engineers solve the golden break? The Harriman Research Tech team is working on it but without support or endorsements development is slower than it should be.
Winner breaks does favor comebacks, but it also favors building a larger lead in the first place. So you can't simply conclude without calculation or modeling which player (or neither) gets the better of it.
I am inclined to model this, but I just got to work. Maybe I'll get a chance to do so later.
This has always been my intuition. The psychological factors of winners break will increase the chances for the weaker player to win compared to alternate/losers break.If you only include two probabilities -- the percentage that A will win against B on A's break and the percentage B will win against A on B's break -- then those percentages give the same match chances for all the break formats. The main theoretical requirement is that the history of the match and the score does not affect those two percentages.
If you include psychological factors somehow -- the players' attitudes, "spine", momentum and so forth -- those may enter into the calculations, but that path is unclear. It is common for fans to say things like, "He gets really good momentum when he starts to put a pack together," and that would be an argument for winner breaks improving that player's chances against someone who doesn't have the "momentum" effect in their game.
The main idea about this in sports is the "hot hand" effect, originally in basketball. The first serious study indicated there was no effect. More recent studies seem to show there may be a small effect in some people. Here's what The Wiki has to say about it:
Hot hand - Wikipedia
en.wikipedia.org
In a race to 7, a 'weaker' player would likely to play much more comfortably and thus better with a 4-0 lead compared to a 4-3 lead, due to psychological factors. All things being equal, the weaker player has a better chance to be up 4-0 in a winners break format compared to alternate break.
I think Matchroom knows this. They also know that the 9-ball format, compared to WPA's 10-ball format, also allows a greater chance for the weaker player to win. I think this is exactly by design. Matchroom wants a greater diversity of players in the final rounds of these tournaments, in order to grow the game.
Last edited:
The nuance can be neverending.
If the gap between the two groups were tightened to say, for example, Ferrari 458 Italias vs ZR1 Corvettes, the Ferraris would still be the "better" group it stands to reason the winner break / alternate break winning % conclusions would change considerably.
Then there's all the other consideerations like Fran pointed out with breaking capabilities, or psychological factors, or maybe did a player have an arguement with their significant other, or equipment issue and on and on with all the variants that can affect play.
Yes, being a better player increases ones chances of winning. Beyond that, it gets murky imo.
Last edited:
Excellent post.
23DenaliBDE
Well-known member
I was trying to figure out a way for chess clock to work. I do think that’s an option if done correctly
Ok. I had a little time and decided to actually simulate it. Obviously this is using math only and assuming that the chance of winning doesn't change based on psychology, fatigue, "hot hand" effect, etc.
https://go.dev/play/p/gtA_WdyXWu5
If you want to run some simulations, just press the run button and it will spit out results. You can adjust the variables of the program; some notes about that:
1. The lag advantage statistics are only valid when the simulation is run with the default lag chance of 0.5 (coin flip). I might some day calculate this in a different way
2. If you adjust the variables and the program completes without producing output, that means you're running too many simulations and the Go playground is shutting it off early. Dial back the number of simulations. The default of 20,000 is a lot already.
Some observations:
https://go.dev/play/p/gtA_WdyXWu5
If you want to run some simulations, just press the run button and it will spit out results. You can adjust the variables of the program; some notes about that:
1. The lag advantage statistics are only valid when the simulation is run with the default lag chance of 0.5 (coin flip). I might some day calculate this in a different way
2. If you adjust the variables and the program completes without producing output, that means you're running too many simulations and the Go playground is shutting it off early. Dial back the number of simulations. The default of 20,000 is a lot already.
Some observations:
- When looking at only match wins, format between winner break, alternate break, and loser break is produces THE SAME RESULTS regardless of how unequal the opponents are and how long the race is.
- As pointed out above, the longer the race, the more likely the better player is to win in the end.
- Although the match win % is the same, alternate break results in closer overall scores, and loser break creates the closest overall scores.
- Winning the lag has a significant impact improving the underdog's chances of winning the match by somewhere around 15%. for a race to 10 with the better player winning the match at ~70/30 clip.
What typical values did you use for "A wins from his break" and "B wins from his break"?
I tried several. It obviously changed the likelihood that either player won the match, but it didn't change the likelihood that they would win with a different format. You can click the link and pick different values. If you find something that produces different results, let me know.
The results are funny and proves my point about stretching stats to prove a argument.
I changed two of your constants and ran the sim a couple of times. First was the race length. I decided to go with something far more common then 10 and opted for 7. Knowing full well that a shorter race is in the lesser player's favour. The second constant was the number of sims. This was cut from the 20,000 down to an actual match, which equals 1.
I ran the above 5 times. In all sims under the winner break format. The stronger player won. Under the Alt break format the stronger player won 3 of 5, and the first two results were the two wins for the weaker player. Now is there any difference between running a sim of 1, 5 times, or sim of 5, 1 time..? No there shouldn't be, but I like to look at things in a case by case basis.
What do I take from this...? Pretty much exactly what I thought.
Lesser player lost all winner break, and won 2 of 5 in Alt break. The rack win percentage was also notably stronger in Alt break for the lesser player. Nothing very surprising here.
In terms of what's going to happen. The stronger player is likely to win, and have a greater chance in winner break. Go figure...
What should also be considered, is that fargo (the current defacto standard for judging player spd) doesn't care about set wins. Only rack wins. So those more closely contested sets of Alt break, will likely hurt the stronger player in rating, and raise the rating of the weaker player.
Your results aren't statically valid because you didn't run enough simulations.
Your last point does expose out a flaw in the Fargo rating system, then.
I see now in the code. I think 0.8 and 0.7 are way high for typical break-and-win chances. I think for all but the very top players, the sum of those two numbers in a match will be close to one. AtLarge has a lot of good data on this. That doesn't change the main conclusion, of course.