How about portable mini tables with 1.25" balls. Whole new market of overpriced professional mini cues.
Fargo on Bar boxes
- Thread starter InsertCleverNameHere
- Start date
I think its harder to play against lower skilled players, not like a 700 playing a 600 but shooting against a weak league player. Its not unusual that they will leave you hooked when they miss because they don't play shape well, they are likely to rearrange the table on you, and they are great at leaving their balls in front of pockets. So almost every time you get to the table against a weaker player there will be problems to solve.
So, this is only an element in winner-breaks formats?
Also, can you address the second part of my question, that is the race calculators having to change based on table size, and now apparently, break format?
At some point, the sample size becomes big enough that you can measure performance accurately. That isn't a fact that is debatable and it applies to everything measureable.
IMHO you have to be way more precise on a barbox table, there are rarely easy runs where you don't have to break at least one cluster, cue ball position needs to be damn near perfect, and unless you do break and run the balls that were open when you started may not be open later in the rack, the cloth is usually slow and beat up, you will be using a magnetic cue ball, and the rails are probably original, the shots are shorter but table conditions are rarely ideal, even at a local tournament.
Sweeping Generalization Ahead
Short races are more of an equalizer than table size. The weaker player can gofrit and stream his 15 minutes worth. In longer sessions especially for cash stakes, the headroom of the stronger player becomes the operative factor. This happens over and over on all equipment.
Short races are more of an equalizer than table size. The weaker player can gofrit and stream his 15 minutes worth. In longer sessions especially for cash stakes, the headroom of the stronger player becomes the operative factor. This happens over and over on all equipment.
InsertCleverNameHere
Well-known member
No matter how unlikely it may be, you are more likely to win a race to 5 against the best player in the tournament on a 7' than on a 9'. Sure, if you make it a race to 50 that removes all doubt but the race would be closer on the 7 than on the 9.
It is an element in loser breaks format as well. It is not explicitly a table size effect. It is about runouts getting too easy. That can happen by some combination of players getting stronger and equipment getting easier. The game is kind of broken at that point when you have too high a fraction of games for which opponent provides no resistance. That’s pool becoming bowling. The race calculator works fine for the vast majority of players/competitions. I think you’d not see any statistically meaningful effect comparing, for example, matches at turning stone (winner breaks) to international open. But you would in a pro level bar box 8-ball tournament.
As Bob mentioned, straight pool is a good example to exaggerate the issue to make it plainly discernible. In straight pool races to 10, the race calculator probably works fine for players under 350 and starts failing sometime after that.
Again, this is a subtlety. The first sentence (more likely to win on 7 foot) is right. For the second sentence it depends what you mean by “race is closer.” If it just means more likely win, then yes. But many think the equipment makes the scores closer, and we don’t see that in actual data. If on average over 100 matches you are 2.2 games behind the best player on the 9’ table, our experience is that’s about where you will be on the 7’ table as well. The scores don’t get closer.
Your initial answer referred to packages, which is only relevant in winner break formats. Otherwise, the other person gets to the table every other game to restart their own (presumably independent) trial.
Bob's example was helpful. It works because each run is its own trial as each shot is not independent--if a person makes a shot (say, they have a 98-99% chance of any one shot) then they are at the table for the next shot. However, I am willing to bet if straight pool were played in some bizarro fashion where each player traded off each shot then the scores would be a lot closer.
Ultimately, your argument that p=.55 but somehow standard deviation changes and break format doesn't matter is unconvincing.
As well, with regards to handicapping, you implied (perhaps just illustratively) that a race to five on a 9' table might be equivalent to a race to seven on a 7' table. Since a lower-rated player would have a better chance of beating a higher ranked player in a shorter race, that means given a fixed race to seven, a worse player would strictly have a better chance on a 7' than a 9'. That means, according to your evaluation of your own system, if one is handicapping two players based on Fargo, that handicapping must depend on table size.
So which table size is Fargo correct for? More likely, Fargo is technically wrong for both since it's an average across all games and all table sizes. The number it's giving us is somewhere in the middle. For example, if Fargo says player A is 60% to win a race to seven against player B, the probabilities are more like (making up numbers here for illustrative purposes) 65% on a 9' table and 56/57% on a barbox.
(I didn't make the difference from 60% uniform under the assumption that for the set of players under 700 Fargo many more games in the system are on barboxes than 9' tables and therefore the given estimate would be closer to the barbox value).
I do appreciate your responses, and your accessibility with discussing your system.
Last edited:
The big table and the bar table require different skill sets.
There are fewer clusters on the big table so long shot making is paramount.
There are more clusters on the bar table so making shots and breaking clusters is paramount.
Bar boxes require more 3-rail shot making to walk CB around interference.
If you play 50%-50% on each, you will develop both sets of skills.
There are fewer clusters on the big table so long shot making is paramount.
There are more clusters on the bar table so making shots and breaking clusters is paramount.
Bar boxes require more 3-rail shot making to walk CB around interference.
If you play 50%-50% on each, you will develop both sets of skills.
InsertCleverNameHere
Well-known member
I agree with this in principle. However at a certain point the ability to strike the object ball accurately you gain playing on the big table translates into being precise enough to navigate tight spaces and clusters. Most players that play well on the big table can take their game to the bar table with ease, doesn't really work the other way around.
Passing the break back and forth doesn't make the trials independent. In a race to 3 between SVB and Gorst playing barbox 8-Ball, winning the lag is really important whether you are doing winner breaks or alternate breaks. There are, in a sense, winner-break packages and alternate break packages.
Yes, agreed.
Ultimately, your argument that p=.55 but somehow standard deviation changes and break format doesn't matter is unconvincing.
Did you watch the couple videos on the FargoRate facebook page about break format and set scores vs game scores?
I am not following your argument here. It is the chance of larger deviations from expectation that is differing, not the expectation itself. If you take a particular handicapped race, say 4-7, the fraction of sets won by the higher player is the same. But you can see more disparate scores (like 4-3 and 0-7) on the easier equipment. On the harder equipment you'll see more like 4-6 and 2-7.
Er...independent trials has a mathematical definition. It is P(AUB) = P(A)*P(B). Any real world competition of course involves a psychological element, but when alternating breaks as in this example, the probability of the person, say Shane, who won the lag winning the third game is the same regardless of the previous games. It is the probability of him winning a rack that he breaks, and therefore is independent from the other racks (trials). In winner breaks, it is dependent on who won the second rack, which itself is dependent on who won the first rack, as it can be the probability he wins when he breaks *or* the probability he wins given Fedor breaks, depending on how the first two racks play out.
The final piece there is that you said that a race to 7 on a barbox might be considered the equivalent of a race to 5 on a 9' in terms of "trials." Well, let's say we have a weaker player vs a stronger player, say me vs SVB. Break format irrelevant. What is more likely--I beat SVB in a single rack or in a race to three? How about a single rack vs a race to 5? How about a single rack vs a race to 7? Or to 120? The longer the race, the more likely it is that SVB wins. Therefore, the chance of my winning a race to 5 is better than the chance of my winning a race to 7. By your own explanations this means I am more likely to win an equivalent-length race on a 7' table than on a 9', given we are essentially comparing a race to 5 to a race to 7. Again, using your explanation.
I'm all for mathematical explanations, but analogies and whatnot, while they can be illustrative, are absolutely not proofs. Your analogies and explanations contradict expectations from basic statistics and mathematics. That is not saying they are necessarily wrong (see: Banach-Tarski paradox) but it would take a mathematical proof and not a comparison with sidewalks to resolve. As of now, I feel that Fargo is a very good general tool, but I do feel like we are fleshing out its limitations and, really, identifying opportunities for more precision/accuracy/granularity in the areas it is used: comparing two players and handicapping races.
Breaking up clusters requires a different skill set, you must know and get proper position to make the object ball AND break the cluster AND leave yourself with proper position to continue the run while navigating this in tight spaces, people who run racks on a bar box have superior cue ball control.
this would be interesting to know.
Someone who has never played a day on a barbox, would immediately have zero issue navigating any aspect of their resultant layouts. It’s like taking candy from a baby.
Yes. For this to be satisfied for SVB and Fedor playing bar-box 8-ball, they's need to be flipping a coin for each break. That would make the events (games) independent. For two games as independent events you have, from SVB's perspective
P(WW) = 0.5 * 0.5 = 0.25
P(LL) = 0.5 * 0.5 = 0.25
P(WL) = 0.5 * 0.5 = 0.25
P(LW) = 0.5 * 0.5 = 0.25
As a result you have 0-2 a quarter of the time, 2-0 a quarter of the time, and 1-1 half the time
For winner breaks, alternate breaks, and loser breaks, the games are not independent events for these runout players. For winner breaks and loser breaks you get almost all 0-2 and 2-0 scores. For alternate breaks you get almost all 1-1 scores.
I see what you are saying that the probability of winning game 3 doesn't depend on the OUTCOME of game 1, But game 1 and game 3 are correlated here. The likelihood is high you either win both or lose both
Yes, this is exactly my point.
I am not sure what you are referring to here. But if you find an example, I will happily concede. I am not wedded to any particular analogy or explanation. I want them to be right and helpful. If they are missing one of those, I will abandon them.