let's talk about averages

[bowl1820]

For a new topic of discussion or at least one that is not talked about as much, let's talk about averages.

For years averages have been used as the main indicator of a players skill level. In the past when someone said they were a league bowler and had a “170” average or the holy grail of averages a “200”. You had a fair idea of how good they were and how you compared to them.

But now with the advent of better bowling balls and high scoring lane conditions. Bowlers scores have risen and thus so have their averages. But these higher averages don't necessarily reflect the players true skill level.

So it may be time to look at changing how averages are figured and/or used to calculate handicap. Or the possibly of having a player skill ranking number used in conjunction with averages for purposes of figuring handicaps and league/tournament or placement in different divisions.

Joe Slowinski wrote a interesting article in the Dec.2007 issue of Bowling This Month magazine. In it he suggested using what he calls the “International Bowling Skill Score Number”.The IBSSNumber would represent a bowlers versatility, accuracy, power, repeatability and knowledge.

You can read a copy here Click Here ForArticle PDF (http://bowlingknowledge.info/images/stories/slowinski_btm_dec_2007_ibssn.pdf)

I can see his idea being used in the international, college and pro bowling areas. But maybe not at the league level of bowling. I think it would be too hard to implement on the league level.

P.S.
Slowinski 's idea is not meant as a replacement for averages, just a way to offer coaches and players a more accurate method of determining a player's strengths and weaknesses and way to compare the ability of bowler's.

[ericj]

As far as league play is concerned, what are you hoping to achieve by grading bowlers differently?

If it's an issue of recognition of relative skill, it's probably unnecessary. I've been bowling for less than six months, but it's already abundantly clear that every bowler in the one league I'm a member of is individually recognized for their specific skills, and there is no obsession with averages.

Except when it comes to handicapping. It is a handicap league. Personally I started with an entirely legitimate average of 83 because i quite literally stank at the game. I joined a few months into the season because an old friend needed another bowler on his team - one of his teammates neglected to mention that he was going to get back surgery in october.

As of the 21st, my official average according to USBC rules was 105. I bowled a 437 series, which was 32 pins better than the previous week. Have i been sandbagging? No! I was a terrible bowler when i started. I'm less terrible now. Everybody recognizes that.

Does it artificially inflate the team score? You betcha.

But i think one of the purposes of the handicap league is to encourage personal improvement. If you want to compete based strictly on your raw score, join a scratch league.

Is it about recognizing the relative skill level based on the equipment used? I think if you wanted to have a league where everyone uses a non-reactive plastic ball with symmetrical weight and traditional drilling with no finger grips, you could do that.

There are certainly still people who throw plastic balls for every frame. Two of 'em on my team. Another is the guy who runs our league.

Should they get extra handicap for that? I don't think they'd appreciate it.

The bowlers that really impress me are some of the unbelievably ancient bowlers on the coed league. iirc the oldest are in their nineties.

They shuffle up to the lane and sort of toss a light-weight plastic ball in what, to the untrained eye, looks like a casual and flippant manner. They can't bend, they can't put any spin on the ball, they just walk up and let it go.

And it works.

[bowl1820]

As far as league play is concerned, what are you hoping to achieve by grading bowlers differently?

What I'm hoping to achieve is to just offer up another area of discussion. The subject is a off shoot of the long running discussion about the integrity of our sport.

The topic of integrity in bowling usually includes talk about Soft Lane conditions and the high tech balls in use today, and how they have allowed players to bowl above their true ability. At least that's how a lot of bowler's see it.

This produces another off shoot topic of discussion which you see a lot in the bowling boards. That is about the handicap system used.

The topic of average calculation comes up rarely compared to the others. Those that bring it up I believe feel that some of the faults in the other areas, May be addressed by having a more accurate measure of a bowlers ability.

My personal thought's are that the average and handicap systems in use now are adequate for leagues and most tournaments.

But I could see having some kind of skill ranking for bowlers. Because you can have bowlers of equal average, but not of equal skill.

Example-
Lets say you have a competition. you have Bowler (A) and Bowler (B) both have a 200 ave. so you put them up against each other. Because they have the same ave. you figure they have equal ability, so it will be a fair competition.

Bowler (A) crushes Bowler (B) every game Why? the games should have been closer, but (A) was a lot better than (B).

Lets look at the how the averages were figured.

Bowler(A)'s average was figured from this
200-200-200=600 =200ave.

Bowler(B)'s average was figured from this
150-150-300=600=200ave.

Which was the better bowler?
(A) who bowled consistent games of 200's or (B) who just just had one high game?

I used a 3 game record for their averages just to show the basic idea of how they could have the same average. but be very different bowler's

Now say if they had been skill ranked in some way, let's say
Bowler(A) 200 ave.Skill Rank #5
Bowler(b) 200 ave.Skill Rank #15

You would know that this match up might not be as fair as you thought and you would match up different bowlers.

As I said something like this would be difficult to implement on the league level. It's just for stimulating ideas and discussion of those ideas.

[kakcpa]

3 games is a pretty small sample size to be using in your example. You can have two bowlers, one that has averaged 225 over the past 3 years, and another that has averaged 220 the past 3 years on a wallyworld condition. Put each of those bowlers on a PBAX condition, and you might see that the guy who averages 220 might completely dominate the 225 bowler...why? The 220 guy may have the talent to hit is mark more consistently, make quicker and better adjustments than the 225 bowler who has an easier time with the walled up condition that allows him 7 boards of lee-way if he makes a mistake.



What I'm hoping to achieve is to just offer up another area of discussion. The subject is a off shoot of the long running discussion about the integrity of our sport.

The topic of integrity in bowling usually includes talk about Soft Lane conditions and the high tech balls in use today, and how they have allowed players to bowl above their true ability. At least that's how a lot of bowler's see it.

This produces another off shoot topic of discussion which you see a lot in the bowling boards. That is about the handicap system used.

The topic of average calculation comes up rarely compared to the others. Those that bring it up I believe feel that some of the faults in the other areas, May be addressed by having a more accurate measure of a bowlers ability.

My personal thought's are that the average and handicap systems in use now are adequate for leagues and most tournaments.

But I could see having some kind of skill ranking for bowlers. Because you can have bowlers of equal average, but not of equal skill.

Example-
Lets say you have a competition. you have Bowler (A) and Bowler (B) both have a 200 ave. so you put them up against each other. Because they have the same ave. you figure they have equal ability, so it will be a fair competition.

Bowler (A) crushes Bowler (B) every game Why? the games should have been closer, but (A) was a lot better than (B).

Lets look at the how the averages were figured.

Bowler(A)'s average was figured from this
200-200-200=600 =200ave.

Bowler(B)'s average was figured from this
150-150-300=600=200ave.

Which was the better bowler?
(A) who bowled consistent games of 200's or (B) who just just had one high game?

I used a 3 game record for their averages just to show the basic idea of how they could have the same average. but be very different bowler's

Now say if they had been skill ranked in some way, let's say
Bowler(A) 200 ave.Skill Rank #5
Bowler(b) 200 ave.Skill Rank #15

You would know that this match up might not be as fair as you thought and you would match up different bowlers.

As I said something like this would be difficult to implement on the league level. It's just for stimulating ideas and discussion of those ideas.

[bowl1820]

3 games is a pretty small sample size to be using in your example. You can have two bowlers, one that has averaged 225 over the past 3 years, and another that has averaged 220 the past 3 years on a wallyworld condition. Put each of those bowlers on a PBAX condition, and you might see that the guy who averages 220 might completely dominate the 225 bowler...why? The 220 guy may have the talent to hit is mark more consistently, make quicker and better adjustments than the 225 bowler who has an easier time with the walled up condition that allows him 7 boards of lee-way if he makes a mistake.

I knew that someone would miss the point and say something about the 3 game record.

I just used 3 games because that's all that was needed to show that the 2 bowlers could have the same average, but bowled differently.

Thats why I said I used a 3 game record for their averages just to show the basic idea of how they could have the same average. but be very different bowler's

I didn't need or want to write out a entire league season of scores. Look at those 3 games as just a "sample from" their league record of how they bowled.

I'll try and put it this way if it makes it easier then
Bowler(A)'s average- he bowled this series 200-200-200=600 every night of league for 34 weeks(102 games) ending average 200

Bowler(B)'s average-he bowled this series 150-150-300=600 every night of league for 34 weeks(102 games) ending average 200

(Now wait for it someone will bring up the 300 games or how consistent they bowled and miss the point too. I just used these numbers because they could be easily and quickly added up. I could have used195-195-210=600 and 175-175-250=600=200ave)

The thing is bowler (A) bowled better more consistent games. Bowler (B) only bowled a few very high games that skewed how high the average was.

I believe in math it has to do with whats called "The standard deviation" which is a statistic that tells you how tightly all the various examples are clustered around the mean in a set of data
bowler (A)'s data is close to the mean (average)
where bowler (B)'s data is at the extremes away from the mean (average).

Now let's forget my examples and use yours. It proves my point too.

You have two bowlers and their averages while they may have shown the average of all their games they bowled. But It didn't represent accurately their true skill level.

[ericj]

There are a lot of statistical models that could be applied to the problem at hand.

My own example, based on my 21-game average that I've been bowling well above, 'proves' your point as well.

When there's money on the line, sure, there would be an obvious advantage to using a better model.

When you're only there for the love of the game, i doubt most people are all that concerned about it.

[The KingPin]

There are a lot of statistical models that could be applied to the problem at hand.

My own example, based on my 21-game average that I've been bowling well above, 'proves' your point as well.

When there's money on the line, sure, there would be an obvious advantage to using a better model.

When you're only there for the love of the game, i doubt most people are all that concerned about it.


Yea I am not sure. If money is involved then yea I can see a lot of things happening. I think it needs to be fair across the board. We should look into this further.

[Jord_84]

basically, this new system of averages would weed out the people who get erratic high games from the people who consistently bowl high games.

I'd be in favor of that.

[Iceman]

Wow - now this is a great topic and nice to see the different views. I agree, the hdcp is good for most leagues and tourneys, but overall, the THS and agressive stock equipment does influence how well (or how much room) they have to acheive their averages.

Like in baseball, to be successful, only 3 out of 10 hits, but in bowling 3 out 10 spares = low average.

I just used that as a starting point. Can't compare baseball to bowling, but in any sport - statistics are influenced in more ways especially when broken down. Another example - not pointing fingers at the steelers :p - but Ben's passer rating is much higher than it should be based on the number of passes = to someone like Palmer per game, not per year or most other QB's. Yards after catch comes into play on QB's total yards, but how much deals on the receiver?

This just shows that all statistics can be "marketed" to show advantage over any or the edge sort of say. Average is average - yards per carry, yards after catch, Runners in scoring position, etc. Even in bowling, 150, 150, 300 and 200, 200, 200 games statistically would make both average players scratch against each other. We all know this is really unfair - but hey, that's where skills come into play. I agree, skill points should be assigned, couldn't they use something like in pool, where rankings are assigned, I'm not too familiar with this ranking system, but I hear it matches ranks against ranks, no matter how much skill or slop shots they get.

Maybe divisions, would balance out the field - 150 and below division, 151 - 170 avg, and each person gets the hdcp based on highest avg within that division, i.e. Bowler with 152 average would get 90% of (170-152) or 16 pins. If someone with a 152 avg goes against someone outside their division avg, they would get 90% of the opponents avg, not their division high. So someone with a 230 avg would give up more pins to the 152 avg, then the current standard 220 hdcp system. This will still only give the 152 avg a few more pins, but any amount of pins would help.

Now, on the 230 avg bowler side, that would seem like a steep hill to climb, but considering the 230 avg person, they would be bowling in the 200's more consistently to achieve that avg, it's impossible to have 2 190 games and still get 690 - so the person with 152 avg, shoots a 150 + his 70 HDCP gives him a 220 game, that's a fair shot for the 230 avg bowler. I'm sure the 230 avg person would complain about this, as why should he give up more pins because he practices to someone who shoots 3 games a week and avg's 150's. Overall, the 230 bowler would win anyway, we all know that even the 230 bowlers have off nights, and the 150 bowlers have nights when everything carries. It happens to all of us.

This is something I'm working on as part of the tournament breakdown and bracket scenario's. Any suggestions?

[ericj]

basically, this new system of averages would weed out the people who get erratic high games from the people who consistently bowl high games.

I'd be in favor of that.

you don't have to get fancy to do that. with enough samples, you can just drop the top 3% and bottom 3% as statistically insignificant.

[bowl1820]

Here was a interesting idea from a post in another forum, where I posed this question regarding averages.

Do you think averages and how they are figured in bowling are a accurate measurement of a bowlers level of skill?


no ... i think there should be some "slope rating" for bowling like there is for golf

you would have to look at all of the averages of bowlers in that house, and how their average in that house compares to their averages in other houses, to determine the "difficulty" of a certain house

but of course to do that, you would have to have some kind of consistency of conditions at each house ... good luck with that at some places

Giving houses a difficulty rating was interesting. I thought maybe some kind of difficulty rating for shots might be factored in to the mix someway.

[JAnderson]

The IBSSNumber would represent a bowlers versatility, accuracy, power, repeatability and knowledge.

While interesting, it is subjective, so beauty is in the eyes of the beholder. For general league play, all of the above traits can actually be counter-productive to scoring. (If you question this, please open another thread and send me a mail!) Unfortunately, scoring is one of the only non-subjective ways to measure performance.

I wish there was a way we could consistently and accurately connect a bowler's scores to their monetary winnings (gross). I'm aware that it is not feasible and will likely never be feasible, but it is a system that could work. A bowler would be rated on his/her scores as compared to monetary earnings. Even "Vegas" leagues "pay out" in the form of a trip to Vegas that has a monetary value. This type of system would, for example, help weed out those who purposely keep low averages in inconsequential money leagues to reap the rewards in high payout tournaments. (Their "average" would be based more on their tournament scores because of the weight afforded)

So just a quick example:

Bowler averages 160 in league and wins $100 for the season
Bowler averages 150 in another league and wins $200 for the season
Bowler averages 200 in a tournament and wins $4700

Bowler's average is as follows

160 x 100 (money won at this average) / 5000 (total money won)
150 x 200 / 5000
200 x 4700 / 5000
----------------------
Add the results for an adjusted average of 197.2

[bowl1820]

While interesting, it is subjective, so beauty is in the eyes of the beholder. For general league play, all of the above traits can actually be counter-productive to scoring. (If you question this, please open another thread and send me a mail!) Unfortunately, scoring is one of the only non-subjective ways to measure performance.


In regard to the IBSSNumber in slowinski's article. I believe it's not meant to be used as a "average" or such in like a league situation. It's just as a bowler ranking system and/or for coaches/bowler's to have a more accurate method of determining their strengths and weaknesses.

As for it being "subjective" Slowinski's four domains: accuracy, versatility, repeatability and power. All can be measured using CAT's and BowlersMAP and DigiTrax according to him (his whole article basically explains how to measure the different parts).

So I'd say you can't say it's subjective, the closest one to being subjective might be the Versatility part.

I don't believe factoring in a bowlers monetary winnings with their average to come up a adjusted average would work. Things like what money would be included? winnings from brackets,pots? what about normal league prize fund payouts would they be included.

If averages are adjusted it should be based on what the bowler bowled.