(Stats) Mosconi Cup: Chris Robinson vs Eklent Kaçi

[PoolStats]

Sure one could justly say that Kaçi may have been lucky getting an easy 2-9 combo and an easy 9-ball for the match clincher in the end. However, of the 5 missed shots by Kaçi, 3 of those were dry breaks and of the 5 shots of Robinson, those were all off the break. Kaçi also made some nice jump shots to get a hard out. If there are any conclusions to make about this match is: Kaçi was unlucky on the break and lucky off of it while Robinson made a few unforced errors and positive luck didn't seem to be on his side.

Notes: Data recorded with the Pool Stats Pro app.

 

[ChrisinNC]

Sure one could justly say that Kaçi may have been lucky getting an easy 2-9 combo and an easy 9-ball for the match clincher in the end. However, of the 5 missed shots by Kaçi, 3 of those were dry breaks and of the 5 shots of Robinson, those were all off the break. Kaçi also made some nice jump shots to get a hard out. If there are any conclusions to make about this match is: Kaçi was unlucky on the break and lucky off of it while Robinson made a few unforced errors and positive luck didn't seem to be on his side.

Notes: Data recorded with the Pool Stats Pro app.

I don’t quite agree with your stats that a dry break should count as a missed shot. I think break statistics, as important as they obviously are, should still be kept separate from shotmaking stats after the break. For instance, what good is it to make 3 balls on the break if you have no shot, as opposed to making one ball on the break and having an ideal layout with a high percentage first shot?

For that matter, safety/defensive shots should also be kept separate. If anything, a poorly executed safety should count the same as a missed shot, and a 2-way shot which even if missed if it leads to a good safety, should not count as a missed shot. It takes a lot of work to compile such detailed stats, but there is certainly some subjectivity and debate involved in completely and accurately evaluating some shots.

 

[spartan]

Sure one could justly say that Kaçi may have been lucky getting an easy 2-9 combo and an easy 9-ball for the match clincher in the end. However, of the 5 missed shots by Kaçi, 3 of those were dry breaks and of the 5 shots of Robinson, those were all off the break. Kaçi also made some nice jump shots to get a hard out. If there are any conclusions to make about this match is: Kaçi was unlucky on the break and lucky off of it while Robinson made a few unforced errors and positive luck didn't seem to be on his side.

Notes: Data recorded with the Pool Stats Pro app.

Nice.
Is Shooting % of match something like TPA?
What is difference between Balls potted and made shots? Balls potted include balls on the break?
What is AVG shot probability?

 

[PoolStats]

I don’t quite agree with your stats that a dry break should count as a missed shot. I think break statistics, as important as they obviously are, should still be kept separate from shotmaking stats after the break. For instance, what good is it to make 3 balls on the break if you have no shot, as opposed to making one ball on the break and having an ideal layout with a high percentage first shot?

For that matter, safety/defensive shots should also be kept separate. If anything, a poorly executed safety should count the same as a missed shot, and a 2-way shot which even if missed if it leads to a good safety, should not count as a missed shot. It takes a lot of work to compile such detailed stats, but there is certainly some subjectivity and debate involved in completely and accurately evaluating some shots.

About dry breaks being counted as a missed shot, I can completely understand the logic behind your concerns. Our reasoning is that a break shot is still, nevertheless, a shot. By our data, players make this shot roughly 73% of the time and if we count it as a made shot, then the converse should also be true. This is why we include the stat Adjusted Shooting Percent, which can be calculated one of three ways. More on that later. To be short about it, this percentage doesn't count dry breaks as a missed shot in it's calculation, therefore adjusting it for people who share a similar concern as you (and we) do.

Safety/defensive shots are kept separate and factor into the calculation of Shooting Percent and adjusted shooting percent. We wrote a paper about the Quasi-subjectivity of pool game data collection. Briefly, what seems subjective at first, is really an educated guess (scientific method?) and through data analysis we can adjust this "subjectivity" to become objective. That's the reason for us coining the term Quasi-subjectivity. If you want to know how effective a player's safety play is, we have a stat for that:

Defensive Success Rate (DSR)
http://poolst.at/dsr

Two-way shots are always calculated in favor of the player. That is, if he/she was attempting to pot a ball, but it is relatively clear by all parties that not making the pot will, most likely, leave an advantageous safety shot, the stat is recorded as such. We have documented all this in previous write-ups that you can find on our website.

I hope that may clear up some of your concerns as there are many to meditated on and we appreciate you asking these as it gives us food for thought.

 

[PoolStats]

Nice.
Is Shooting % of match something like TPA?
What is difference between Balls potted and made shots? Balls potted include balls on the break?
What is AVG shot probability?

Shooting Percent is very similar to TPA in fact. In recording Accu-Stats games we've noticed a slight variation in between our SP% and Accu-Stats TPA - somewhere on the measure of 1-1.5% if you adjust TPA to our metric (100*TPA). There are again three ways to calculate this: 1. Where defensive shots hurt the calculation (bad idea!) 2. Where defensive shots are neutral (fair, but underrepresented). 3. Where safety play positively effects SP% (ours). However, 3. begs the question, should a bad safety play produce a positive outcome in SP%? A fair question, but there are many situations where a player is locked up and is solely trying to make contact with the object ball leaving a high likelihood that they will leave something advantageous to their opponent. Also, no stat is 100% accurate, that's why it's called a stat. But, if the concern is that bad safety play should affect any calculation negatively, then we refer you to our stat Defensive Success Rate which does just this.

A player may pot 3 balls on the break and shoot a combo on the next shot making two balls in. So, there were 5 balls potted and only 2 made shots.

AVG Shot Probability is a Quasi-Subjective stat. It's an educated guess on the level of difficulty for any given shot. We preface it with quasi, because we can analyze the data after and adjust this Shot Probability in accordance with the trend of the population per what percent of people make what difficulty of shot. That is if we record a certain type of shot as a probability of .7 (or 70% chance), but find out that the population of players make these .7 shots with a .814 probability (81.4% chance), we can adjust the data in accordance with what the population of the sample represents. These adjustments will be made over and over again in our data which will refine it more and more eventually hitting a limit point.