Break Stats -- 2017 U.S. Open 9-Ball Championship, Oct. 2017

Here are some aggregate break statistics from the 39 9-Ball matches streamed this past week on a pay-per-view basis by Accu-Stats from the 2017 U.S. Open 9-Ball Championship at the Sheraton Norfolk Waterside Hotel in Norfolk, VA. Jayson Shaw won the event.

Conditions -- The conditions for these streamed matches included:

The 39 matches (676 games) streamed were as follows, shown in the order in which they were played. These 39 matches represented 14.2% of the event's total of 274 matches played (excludes the 2 forfeited matches). The figures in parentheses for some of the matches are the Accu-Stats Total Performance Averages (TPA), as calculated by Accu-Stats and shown on the stream.

[Note: These stats exclude 7 games that I missed -- 6 in the Ouschan/Chinakhov match and 1 in the He/Orcollo match. So the total number of games tracked was 669.]

Sun., Oct. 22, 2017

Mon., Oct. 23

Tues., Oct. 24

Wed., Oct. 25

Thurs., Oct. 26

Fri., Oct. 27

Sat., Oct. 28

Overall results -- The breaker made at least one ball (and did not break illegally or foul) 62% of the time (417 of 669), won 57% of the games (380 of 669), and broke and ran 26% of the games (176 of 669). The break was illegal 10% of the time (66 of 669).

Here's a more detailed breakdown of the 669 breaks.


Break-and-run games -- The 176 break-and-run games represented 26% of all 669 games, 46% of the 380 games won by the breaker, and 42% of the 417 games in which the break was successful (made a ball, legal, no foul).

The 176 break-and-run games consisted of 1 6-pack (Van Boening), 6 4-packs (2 by Sanchez Ruiz and 1 each by Morris, Schmidt, Shaw, and Feijen), 5 3-packs, 17 2-packs, and 97 singles.

9-balls on the break -- The 176 break-and-run games included 11 9-balls on the break (1.6% of the 669 breaks). One additional 9-ball was made on the break when the breaker scratched, so it was spotted. With the Accu-Rack, the 9-ball tends to remain close to its original position.

Miscellany from the data for the 2017 U. S. Open 9-Ball Championship
[This relates only to the 39 streamed matches, not to all matches in the event.]

• The most balls made on a single break was 4, once by Melling (a B&R) and once by Van Boening (including the 9-ball).

• The average number of balls made on all breaks was 1.1. On successful breaks (legal, made at least one ball, and did not foul), the average was 1.5.

• 49% (329 of 669) of the games ended in one inning – 26% (176) won by the breaker (B&R) and 23% (153) won by the non-breaker. 9% (60 of 669) of the games lasted more than 3 innings.

• 40% (265 of 669) of the games were run out by the player who was at the table following the break. These run-outs were:

• The player who made the first ball after the break:

• For the 38 races to 11 (i.e., excluding the finals race to 13), the loser won an average of 6.3 games. Four matches went to hill/hill; one was a shutout.

• The longest match in elapsed time, at 163 minutes, was Kaçi d. Appleton 11-8. The shortest match in elapsed time, at 57 minutes, was Kiamco d. Mscisz 11-0. The elapsed time was measured from the lag until the winning ball was made (or conceded), so it includes time for racking and timeouts.

• The average elapsed time for the 38 races to 11 was 105 minutes.

• The average minutes per game for all 39 matches was 6.0.

• The two matches highest in average minutes per game, at 8.6 min./game, were Kaçi d. Appleton 11-8 and Kaçi d. Sanchez Ruiz 11-7. The match lowest in average minutes per game, at 4.1 min./game, was Schmidt d. Cicero 11-3.

• Breaking fouls averaged about 1 for every 11 games, other fouls 1 for every 6 games, and missed shots 1 for every 2.2 games.

• About 35% of the games involved one or more safeties.

Here are some comparisons for the US Open 9-Ball streamed matches I watched for this year versus the last two years.

Breaks on which at least one ball was pocketed (includes all wet breaks even if illegal or fouled)
2017 -- 71% (475 of 669)
2016 -- 73% (306 of 422)
2015 -- 68% (388 of 572)

Illegal breaks (whether wet, dry, or fouled)
2017 -- 10% (66 of 669)
2016 -- 20% (86 of 422)
2015 -- 4% (23 of 572)

Successful breaks (broke legally, made at least one ball, and did not foul):
2017 -- 62% (417 of 669)
2016 -- 57% (240 of 422)
2015 -- 62% (353 of 572)

Breaker won game -- on all breaks:
2017 -- 57% (380 of 669)
2016 -- 54% (227 of 422)
2015 -- 53% (302 of 572)

Breaker won game -- on legal breaks:
2017 -- 57% (345 of 603)
2016 -- 57% (191 of 336)
2015 -- 54% (294 of 549)

Breaker won game -- on illegal breaks:
2017 -- 53% (35 of 66)
2016 -- 42% (36 of 86)
2015 -- 35% (8 of 23)

Break-and-run games -- on all breaks:
2017 -- 26% (176 of 669)
2016 -- 23% (96 of 422)
2015 -- 23% (131 of 572)

Break-and-run games -- on successful breaks:
2017 -- 42% (176 of 417)
2016 -- 40% (96 of 240)
2015 -- 37% (131 of 353)

Average number of balls made on the break:
2017 -- 1.1 on all breaks, 1.5 on successful breaks
2016 -- 1.1 on all breaks, 1.6 on successful breaks
2015 -- 0.9 on all breaks, 1.4 on successful breaks

Games ending in one inning:
2017 -- 49% (329 of 669)
2016 -- 49% (205 of 422)
2015 -- 45% (258 of 572)

Run-outs from first shot after break:
2017 -- 40% (265 of 669)
2016 -- 40% (170 of 422)
2015 -- 36% (204 of 572)

Games won by player who made the first ball after the break:
2017 -- 77% (508 of 658)
2016 -- 76% (319 of 420)
2015 -- 75% (427 of 570)

Average minutes per game (includes time for racking and timeouts):
2017 -- 6.0
2016 -- 6.0
2015 -- 6.1

BRussell said:

... One thing wasn’t clear to me: If it was an illegal break, there was no push-out option for the breaker if the opponent gave it back, is that right?

Click to expand...

Right, neither player was allowed to push out after an illegal break.

That is different from the Euro Tour / WPA rule under which the non-breaker cannot push out if he takes the table, but the breaker can still push out if the table is returned to him.

As happened last year, the percentage of breaks that were illegal changed a lot this year.

In 2015, the 3-point rule used the side pockets for measuring whether the break was legal, and illegal breaks were just 4% of the total.

In 2016, the rule was changed to using the head string, and balls had to pass the head string (base of ball) to be counted. The result was that illegal breaks jumped enormously to 20% of all breaks.

For 2017, the head string was still used for the determination, but a ball only needed to touch it to be counted. I.e., if the edge of a ball touched the vertical plane rising from the head string, it counted. The result was that illegal breaks were cut in half -- 10% of all breaks.

So, for 2017, the measuring line was essentially moved half a ball's width closer to the rack. It doesn't seem logical to me that moving that line by just 1 1/8" should drastically change the percentage of illegal breaks.

Illegal breaks dropped from 14% in 2016 to 4% in 2017 for breaks that were not dry or fouled. If that drop had not occurred, I estimate that the overall break-and-run percentage would have been just about the same this year as last year instead of rising by 3 points.

Cardigan Kid said:

At large, can you isolate just the ko brothers break stats, successful vs unsuccessful
And compare to just Jayson's break stats?

With the racking controversy in the big Ko match, I watched Jayson appear to express dissatisfaction with a number of the racks and break anyways, but harder than Ko...which lead to ko getting more unsuccessful breaks. It wasn't as if Jayson was getting better racks.

Little Ko appeared to suffer the same fate, in regards to lack of power in his cut break.

I'm going out on a limb and say it had nothing to do with 1 or 2 racks.

Do the stats back me up?

Click to expand...

Yes, I can do that. I was planning to do stats for Shaw and Kaçi vs. the rest of the field. They had the most appearances on stream, and were the finalists. But I'll also look at the Ko's. Must do something else now, though, so that will be later.

AtLarge said:

Cardigan Kid said:

At large, can you isolate just the ko brothers break stats, successful vs unsuccessful
And compare to just Jayson's break stats?

With the racking controversy in the big Ko match, I watched Jayson appear to express dissatisfaction with a number of the racks and break anyways, but harder than Ko...which lead to ko getting more unsuccessful breaks. It wasn't as if Jayson was getting better racks.

Little Ko appeared to suffer the same fate, in regards to lack of power in his cut break.

I'm going out on a limb and say it had nothing to do with 1 or 2 racks.

Do the stats back me up?

Click to expand...

Yes, I can do that. I was planning to do stats for Shaw and Kaçi vs. the rest of the field. They had the most appearances on stream, and were the finalists. But I'll also look at the Ko's. Must do something else now, though, so that will be later.

Click to expand...

The 39 matches involved 45 different players. 26 of them appeared just once, 13 appeared twice, 1 appeared three times, 3 four times (Thorpe, Ko Ping-Chung, and Sanchez Ruiz), 1 five times (Shaw), and 1 six times (Kaçi). Here are some stats for each of the 5 players who appeared at least four times.

Successful breaks (broke legally, made at least one ball, and did not foul):

Breaker won the game:

Break-and-run games, on all breaks:

Break-and-run games, on successful breaks:

As for Ko Pin-Yi, he appeared on stream only twice, with 22 breaks. He was successful on 17 breaks (77%) and won 16 of those 22 games (73%) including 6 by B&R (27%). In his match with Shaw, Ko had 7 successful breaks (won 3, lost 4), 2 dry breaks (won both), and 2 wet-but-illegal breaks (won one). He also won 4 games on Shaw's break, for his total of 10 wins.

vapoolplayer said:

That’s the beauty of actually measuring things statistically. Many things that seem “logical” are really just the illusion of logic.

It will take many more matches of the same format to prove, but my theory is you have stumbled upon the area where the averag person’s break power has the balls slowing down. It has to happen somewhere on the table.

The balls are racked in the same formation and thus will go in similar directions if broken from the same cue ball location with close to the same velocity. Obviously there is going to be a lot of variance in location and velocity. But there is going to be an average break speed across the board and balls will on average start slowing down about the same time.

Thus far, the head string seems to be ther area where things slow down for balls that head up table. So if you have found the general line where this happens, any movements to this line can greatly change things. Even if it’s just half of a ball diameter.

Again, this is just me talking out of my ass. You’d have to do thousands or millions of break analysis to sort it out.

Click to expand...

Instead of using the word "logical," I probably should have said it was my "expectation" that moving the line half a ball wouldn't make much of a difference. But the numbers did come out quite different, and moving the line that small amount may have been, as you theorized, the major reason for the change. But other factors may also have contributed to the change. It was a different group of players on a different table under possibly different weather conditions, etc. It's possible that many players this year had a heightened knowledge of how to break legally under the 3-point rule. And, as you note, the numbers are still pretty small here, so fluctuations can be expected.

Thanks for contributing.

Bob Jewett said:

Perhaps you should also note the Accu-stats clock wrinkle: if the player is down on the shot and the clock expires it is not a clock violation as long as the player shoots before getting up.

Click to expand...

I added it, Bob. Thanks.

Here's a day-by-day listing of some break results of the 2017 U.S. Open 9-Ball Championship. The stats are for the 39 streamed matches (excluding 7 games, as mentioned in post #1) -- 6 matches on each of the first 6 days and 3 matches on Day 7.

Successful breaks -- breaker broke legally, made at least one ball, and did not foul:

Breaker won the game:

Break-and-run games -- on all breaks:

Break-and-run games -- on successful breaks:

Total Performance Averages

Accu-Stats calculated and posted final TPAs for 25 of the streamed matches (see post #1 above). Some tidbits:

• The distribution of TPAs was

• Jayson Shaw had the highest TPA, .971. He had 4 of the 12 scores in the 900's. The arithmetic mean of his 5 TPAs was .923.

• TPAs were given for 4 of runnerup Kaçi's matches. He had a high of .913 and a mean of .877.

• Arithmetic means for the 25 matches -- .890 for match winners, .835 for match losers, .863 combined.

• Medians for the 25 matches -- .892 for match winners, .847 for match losers, and .8695 combined.