Calculating BPI

[stevekur1]

I've been keeping track of my BPI during 14.1 League matches. Both with safeties and without safety innings. And I believe my average BPI isn't that bad !

So my question arises that what is the proper way to keep track of BPI ?

And in a 150 point game what would a good average BPI be and how would you rate a player based on his/her BPI ?

Thanks
-Steve

 

[Double-Dave]

BPI is calculated including safeties.

As a rule of thumb, anything over 10 (in the long run, not just 1 match) is considered excellent,
in Germany that would be about what you need to have a winning percentage in the 1st Bundesliga
and those guys shoot pretty straight, we're talking guys like Feijen, Souquet, Roschowsky, vd Berg.

gr. Dave

 

[stevekur1]

BPI is calculated including safeties.

As a rule of thumb, anything over 10 (in the long run, not just 1 match) is considered excellent,
in Germany that would be about what you need to have a winning percentage in the 1st Bundesliga
and those guys shoot pretty straight, we're talking guys like Feijen, Souquet, Roschowsky, vd Berg.

gr. Dave

With safeties and fouls both intentional and on err I've been averaging between a 6 and 7 so I guess that not half bad considering the names that are attached to a 10 BPI

thanks
-Steve

 

[sjm]

There is no such thing as BPI without safeties.

Here are the balls per inning figures from the US Open 14.1 event of 1967.

1st place finisher Jimmy Caras had an 11.02 BPI
2nd place Luther Lassiter had 6.86
3rd place Irving Crane had 9.29
4th place Dallas West had 8.26
5th place John Ervolino had 7.00
6th place Jack Breit had 6.66
7th place Joe Balsis had 10.70
8th place Frank McGown had 6.86
9th place Steve Mizerak had 9.23
10th place Dan Gartner had 8.16
11th place Alton Whitlow had 6.66
12th place Maynard Parish had 4.81
13th place Lou Butera had 10.05
14th place Cisero Murphy had 7.96

Mosconi, reportedly, typically had a BPI of 15 in his prime and Mike Sigel 13.

I have discussed the subject of how many ball one must run on an open table inning with some players over the years. To even contend for the big pro titles in 14.1, the great Jack Colavita reckoned you needed to average 35.

 

[14-1StraightMan]

Safeties should not count.

Yes, counting safeties has always been the way BPI has been totaled but that does not mean it is the correct way. Sometimes changes are necessary. Example:..... Player A comes to the table after a opening break shot (which is a safe by Player B) and runs 75 balls & then plays a safe. That is one inning for that player A.
Now, both players get into a safe match & 20 safes goes back & worth.
The score is still 75 -0. Player A gets up & calls a shot & pockets it & goes on to run 75, game over. 150-0. Yes, it was a 22 total inning game but only two shot innings were attempt to pocket a ball. Why should 20 innings count against a player's average that they did not attempt to make a ball? Answer: It should not. So, that player has a two inning point scoring game.
The only thing that I can compare it to is baseball. When a batter gets up and Walks..... A walk does not count in the batter's average. That is the same thing as a Safe.

 

[dmgwalsh]

Bpi

With safeties and fouls both intentional and on err I've been averaging between a 6 and 7 so I guess that not half bad considering the names that are attached to a 10 BPI

thanks
-Steve

Steve: Between 6 and 7 is pretty solid.

In our leagues, I unfortunately usually reside in the 3-3 1/2 area. Lately, I have been pushing up toward the 5 bpi area and I feel pretty good about my game.

Mike: I know bpi sometimes seems not to tell you much when there are a lot of safes or few good opportunities to shoot, but that's just the way it is. Have to either assume things will average out in the long run, or take the number with a grain of salt.

 

[AtLarge]

There is no such thing as BPI without safeties. ...

The existence of one type of statistic does not preclude the determination of, and discussion about, another stat. Other sports have added to the list of stats many times.

Yes, traditionally, "balls per inning" was the reported stat. But, as other posters have noted, that stat gives quite a limited picture of what happened in the game. It's also a misnomer, because the numerator is points scored rather than balls pocketed. For example, a player may have made 175 balls to end up with 150 points.

Why not look at it in more than one way -- with and without counting safeties and intentional fouls in the denominator? When I reported stats for the Accu-Stats "Make-It-Happen" 14.1 Invitational 5 months ago, I calculated two scoring averages for each player for each game. Here's what I said in that thread:

"In the following stats, the term Attempted Scoring Innings means the player's total number of innings for the game minus the number of innings that he played only a safety or an intentional foul (no attempt to score a point). PPI is Points per Inning based on the total number of innings in the game for that player. [Note: I am calling this measure Points per Inning rather than Balls per Inning to be a bit more precise in the definition.] PPASI is Points per Attempted Scoring Inning, i.e., points per inning based on the number of Attempted Scoring Innings in the game for that player."

Obviously, the two measures can produce quite different results. Here is that thread again: http://forums.azbilliards.com/showthread.php?t=330023

[Note: A third measure is Points per Scoring Inning, which would be total points divided by the number of innings in which points were scored. I preferred to also count innings that consisted simply of one missed shot, hence my Points per Attempted Scoring Inning.]

 

[sjm]

The existence of one type of statistic does not preclude the determination of, and discussion about, another stat. Other sports have added to the list of stats many times.

Yes, traditionally, "balls per inning" was the reported stat. But, as other posters have noted, that stat gives quite a limited picture of what happened in the game. It's also a misnomer, because the numerator is points scored rather than balls pocketed. For example, a player may have made 175 balls to end up with 150 points.

Why not look at it in more than one way -- with and without counting safeties and intentional fouls in the denominator? When I reported stats for the Accu-Stats "Make-It-Happen" 14.1 Invitational 5 months ago, I calculated two scoring averages for each player for each game. Here's what I said in that thread:

"In the following stats, the term Attempted Scoring Innings means the player's total number of innings for the game minus the number of innings that he played only a safety or an intentional foul (no attempt to score a point). PPI is Points per Inning based on the total number of innings in the game for that player. [Note: I am calling this measure Points per Inning rather than Balls per Inning to be a bit more precise in the definition.] PPASI is Points per Attempted Scoring Inning, i.e., points per inning based on the number of Attempted Scoring Innings in the game for that player."

Obviously, the two measures can produce quite different results. Here is that thread again: http://forums.azbilliards.com/showthread.php?t=330023

[Note: A third measure is Points per Scoring Inning, which would be total points divided by the number of innings in which points were scored. I preferred to also count innings that consisted simply of one missed shot, hence my Points per Attempted Scoring Inning.]

I'm in favor of the development of such a stat, but it needs discussion.

For example, would you consider an inning that consists of one or two balls pocketed followed by a safe? Sometimes, it happens because a player failed to make shape onto a possible break shot, but sometimes it occurs because no break shot was realistically available.

I'd advocate that some subjectivity is called for, and that the stats keeper be allowed to judge whether each inning is offense or defense oriented, and the stat I'd approve of is points per offense oriented inning rather than points per attempted scoring inning.

Still, BPI is what permits comparison of today's players with those of yesteryear, and that makes it more interesting to me personally.

Thanks for sharing your thoughts.

 

[AtLarge]

I'm in favor of the development of such a stat, but it needs discussion.

For example, would you consider an inning that consists of one or two balls pocketed followed by a safe? Sometimes, it happens because a player failed to make shape onto a possible break shot, but sometimes it occurs because no break shot was realistically available.

Yes, sometimes the player's intent at the start of an inning is simply to make a ball or two and play safe, not to attempt to run many balls. But my Points per Attempted Scoring Inning counts (in the denominator) every inning in which the player made at least one ball or attempted to do so.

I'd advocate that some subjectivity is called for, and that the stats keeper be allowed to judge whether each inning is offense or defense oriented, and the stat I'd approve of is points per offense oriented inning rather than points per attempted scoring inning. ...

That's certainly possible to do, and some stats in other sports do involve judgment by the scorekeeper. But I may just stick with pure objectivity in what I do. We obviously don't really know what was in the player's mind at the start of an inning, and his plans and actions can change with every shot.

Thanks for the comments on this, sjm.

 

[iusedtoberich]

What about instead of balls per inning, you did ball pocketing probability?

This would exclude safeties, so careful gameplay counting would be required. The calculation would be:

# of balls pocketed in a game
Divided by
# of offensive shots attempted

I use this method quite successfully and I have a running 6 month count of my ball pocketing probability. However, it's very easy for me to keep track of because I am only playing against myself. Thus, every shot I take is an offensive shot attempt.

 

[AtLarge]

What about instead of balls per inning, you did ball pocketing probability?

This would exclude safeties, so careful gameplay counting would be required. The calculation would be:

# of balls pocketed in a game
Divided by
# of offensive shots attempted

I use this method quite successfully and I have a running 6 month count of my ball pocketing probability. However, it's very easy for me to keep track of because I am only playing against myself. Thus, every shot I take is an offensive shot attempt.

Sure, you could do something like that, but it would have a different meaning from the points-per-inning measures. Seems like you would want to exclude intentional fouls as well as safeties. And I wouldn't call it a probability, since it would represent past performance rather than a likelihood for the future. Actually, it could essentially be an Accu-Stats performance average for 14.1: total balls pocketed divided by the sum of balls pocketed and errors (we'd have to define all possible errors).

 

[iusedtoberich]

Yes, very similar concept to accustats rating.

Yes, it could not be compared to balls per inning.

But, it could be a useful comparison tool for measuring ones own improvement, or lack there of.

 

[TSW]

Accu-stats or pocketing probability are more useful measures IMO.

Balls per inning doesn't tell much because it doesn't tell how the inning ends. Obviously, 9 good safes followed by a run of 50 is a very different class than running 5 balls and missing on an open table, even though they both record a BPI of 5.