Is your handicap league fair?

[swingset]

No handicaps are fair. Giving pins that aren't earned is never fair.

[bltuneup]

Okay loaded up the trial version of Bowling League Tuneup.

Pros:
It has a nice clean interface, entering information is pretty straight forward.

Cons so far:
It requires you use Microsoft Excel.
(That's a deal breaker, I and many others use Apache OpenOffice (Which handles Excel files just fine, but doesn't appear to work with BLT.) or other spreadsheet programs.

Which I assume is what caused the error messages I received:
When clicking the Analyze button: BLT2.blt File not found or Contains invalid Data (The file is there in the directory, so it must be invalid data.)

When trying to save league: Error saving league file: ActiveX component can't create document.

Program is limited to 24 teams (We have several full house leagues of 32 teams)

While maybe not a major concern, it can't factor in the Match Point System. Which is a common league format.

Suggestion:
Program should include a sample league file (or a download link for one)

Bummer! Well, I genuinely appreciate your trying it out. Yes, those errors indicate my app desperately wants genuine Microsoft Excel and not an imitation. I really didn't want to require it, but it saved me from a hellacious amount of coding and the use of a separate database. I'm using Microsoft's Excel object library, so I guess it's not a huge surprise it doesn't work for non-Microsoft spreadsheet software, even when it’s in Excel mode.

Your other comments are quite valid. I apologize in the user manual for limiting league size to 24 teams. User interface issues became too hard for me to deal with beyond that, so I reluctantly set that limit. I also mentioned in the manual there were league formats that weren't handled yet, but could be in the future. I had to limit my feature set lest I be coding the system for 10 years before releasing it! The good news is that it shouldn't be hard to adapt my code for other formats in the future.

You’re in agreement with a friend of mine who said I need to include some sample files (a la the Northwind database that comes as a sample file in Microsoft Access). You're both right, but just adding some sample files to what I distributed with the app wouldn't be enough. I'd then need to have a section in the manual to guide a new user through the use of those files. Again, I opted to draw the line somewhere so I could make an initial release in this lifetime. But your point is very well taken, and in the next release I think I'll rewrite the Quick Start section of the manual to include loading and working with the sample files.

Not that I'm anxious to criticize my own app, but you didn’t mention what I feel is the app's biggest shortcoming. You can get so much more out of it if you input bowlers’ league scores, but typing scores in manually can be very time-consuming. So the app really needs an import feature. From what I can gather, CDE Software’s BLS system seems to be widely used, so I'm in the process of trying to contact them in Seattle to find out what it would take for me to be able to export bowler data from BLS.

Again, thank you for taking the time to look at the app. This kind of feedback means a lot to me!

[Mike White]

Just curious, you've mentioned how your team "finished near or at the bottom of the standings for 10 straight seasons".

What handicap was your league using then? and how many teams did it have?

What was your team average then? and what were the averages of the teams that won?

When your ran the numbers for your league and it showed your company league was grossly unfair.
What did it say would have been a fair handicap for your league?

The $64,000 question, unfortunately so far hasn't been answered.

To me, for a handicap league to be "fair" from top to bottom in averages, there needs to be limits on teams entering averages.

If a league consists of mostly 160-180 4 person teams, you might want to limit each teams entering average to 700, or 850 based of 4 or 5 members, and suggest the minimum team entering average be 650, or 800. That way no team can enter above the limit, and teams entering below the minimum accept the "unfairness" ahead of time.

If rules like there were common, I think there might be a revival in scratch leagues.

One handicap league I bowled in recently shocked me as to the "stacked" teams.

My team consisted of 190, 160, 210, 210 average bowlers, and the opposing team was giving us 70+ pins of handicap.

[Aslan]

I must say, Mike's comment is spot on.

I've thought about creating a sport league where it's scratch but there are caps on the team averages. Thats really the only way a scratch league can be truly successful outside of Vegas. Sure, there are scratch leagues that exist...but they are few and far between and if you're not average 200+ there's really no point to it.

But, if you did a 3-man scratch, sport league...make 525 the cap...suddenly the league can be very competitive without any need for handicap (except for the occasional bracket and/or sidepot).

[bltuneup]

The $64,000 question, unfortunately so far hasn't been answered.

To me, for a handicap league to be "fair" from top to bottom in averages, there needs to be limits on teams entering averages.

If a league consists of mostly 160-180 4 person teams, you might want to limit each teams entering average to 700, or 850 based of 4 or 5 members, and suggest the minimum team entering average be 650, or 800. That way no team can enter above the limit, and teams entering below the minimum accept the "unfairness" ahead of time.

If rules like there were common, I think there might be a revival in scratch leagues.

One handicap league I bowled in recently shocked me as to the "stacked" teams.

My team consisted of 190, 160, 210, 210 average bowlers, and the opposing team was giving us 70+ pins of handicap.

Mike, it can certainly be helpful to cap team averages. I begged for that in our company league to prevent the team of four bowlers all averaging over 200 from dominating our league. The argument I got back was that people like bowling in the league because they’re bowling with the people they work with. Capping team averages would force teams to split up, at which point many would lose interest in being in the league. So much for that idea.

But even limiting team averages doesn’t solve the fairness problem…it can just make a league a little less unfair by preventing the creation of a full team of ringers. If some teams come in at the team maximum average and others come in way below that, you still have a fairness problem. If, on the other hand, people treat that maximum average as a goal to shoot for when assembling their teams and all teams therefore come in close to that average, fairness is no longer an issue. Such a league could even bowl scratch and all would be fine.

The big problem with setting team limits is the cheating that can go on to allow a team of ringers to get themselves under that limit. Determining team eligibility based on first-week scores is a recipe for disaster, as the good teams will have huge incentives to keep their scores down that week. Eligibility must be based on book averages. But any bowler without a sanctioned book average would still be able to cheat the system by deliberately bowling poorly to start the season. I really don’t like any rule that incentivizes people to cheat.

And in many of my capped leagues, the ultimate winner achieved an average above the cap, leading to a lot of grumbling and bad feelings among the other teams.

So I’ve come around to the notion that it’s better to give all teams a fair shot not by setting rules to try to create a league that will work well with the chosen handicap formula, but rather by setting a handicap formula that is tailored to the attributes of a specific league and the exact mix of teams in the league. My app lets you do that.

[Aslan]

Mike, it can certainly be helpful to cap team averages. I begged for that in our company league to prevent the team of four bowlers all averaging over 200 from dominating our league. The argument I got back was that people like bowling in the league because they’re bowling with the people they work with. Capping team averages would force teams to split up, at which point many would lose interest in being in the league. So much for that idea.

There's the rub. I've proposed multiple ways to improve league bowling...and most of them involve "inconveniencing" people...which everyone immediately says is a bad idea.

But...league bowling also has been dieing...and much of that is the result of making it "convenient" rather than worry about it being fair.

Points/Examples:
- > 175 average bowlers have no business in a no-tap league.
- > 190 average bowlers have no business in a handicap league.

Fix those two problems and then cap team averages and force any entering person to start with the league average on Day 1 or use their most recent USBC average (if they have one in the system)...and you've essentially solved the NEED for handicap.

[bltuneup]

Just curious, you've mentioned how your team "finished near or at the bottom of the standings for 10 straight seasons".

What handicap was your league using then? and how many teams did it have?

What was your team average then? and what were the averages of the teams that won?

When your ran the numbers for your league and it showed your company league was grossly unfair.
What did it say would have been a fair handicap for your league?

I had to go back through my league sheets to find the data for my company leagues. We had four-bowler teams, with league size varying from eight teams (summer leagues) to 16 teams. League lengths varied from 12 weeks (summer leagues) to 33 weeks. A 90% handicap percentage was used every season, with adjustments occasionally made to the base number in a (sometimes unsuccessful) attempt to keep it above the best bowler’s average.

The leagues set no caps on team averages, which created a large disparity between the best and the poorest teams. I’ll use my league sheet for the fall 2006 season. It was a 33-week league with 11 teams, using a handicap formula of 90% of 210 (which was greater than the best bowler’s average). The season ended with the teams having the following scratch averages (best-to-worst): 733, 708, 688, 665, 640, 631, 614, 598, 587, 561, 558. (I was on the 587 team. I ended with a 185 average, but my teammates obviously didn’t do nearly as well.)

My app didn’t exist back then, but running the numbers retroactively, here are the chances of a first-place finish for each of those 11 teams, based on their final team averages: 34.8%, 23.3%, 16.3%, 10.3%, 5.9%, 5.1%, 3.4%, 2.5%, 1.6%, 0.8%, 0.7%. Trophies were given to second-place finishers as well, so here are the chances of a team getting a trophy (derived from adding the chances of a first-place finish and a second-place finish): 55.4%, 42.8%, 33.2%, 23.4%, 14.6%, 12.9%, 9.1%, 6.7%, 5.2%, 2.9%, 1.6%.

Right off the bat, what strikes me (pardon the pun) is that one of the 11 teams is actually more likely to get a trophy than not. That just seems so wrong. (Oh, did I mention? That team ended up in first place.)

But before we declare this league to give the best teams too much of an advantage, we need to determine what kind of performance it might take for the poorest team to put itself into contention. Let’s assume the averages I gave are the ones the teams entered the season with, and let’s assume everyone on the poorest team was able to raise their averages five pins from these averages, adding 20 pins to the team average. This would be a tremendous team accomplishment. Let’s further assume that every bowler was able to average two pins over the average used to set their handicaps over the course of the 33-week season. This is the new measurement my app introduces, which I call a trend. The longer the season, the harder it is to maintain a very high (or very low) trend. Using tables I’ve published in my user manual, a team could be expected to achieve a cumulative trend of +8 in a 33-week season less than one season in 20. So a season in which a team both raised their average 20 pins and registered a trend of +8 would be a banner season for that team.

Running the numbers after making these assumptions (while keeping all other teams at their expected levels of performance) produces the following chances of a team getting a trophy: 53.2%, 40.3%, 30.4%, 21.6%, 13.6%, 11.6%, 8.2%, 6.0%, 4.5%, 2.4%, 17.8%. So yes, the poorest team’s chances went way up. But there are still four teams with better chances of getting a trophy, with the best team still three times more likely to get a trophy for bowling a typical season than the poorest team is for bowling the best season it’s ever likely to have.

As we’ve discussed, fairness is subjective. But I would strongly disagree with anyone who looks at these numbers and says this league is fair!

At this point, you’d try different handicap percentages to try to achieve a more competitive balance for this league. Going to 95% while continuing to assume a breakout season for the poorest team, we get these chances of getting a trophy: 33.6%, 27.8%, 23.8%, 19.9%, 16.4%, 14.5%, 12.3%, 10.5%, 9.6%, 7.2%, 34.6%. So this setting means that the best team bowling a typical season has the same chance of a trophy as the poorest team having a great season. That certainly sounds fairer to me!

Is this the best percentage? That’s where subjectivity comes in. Some might say there’s still too strong a bias toward the better teams. And should a team having the season of its life still only have one-third of a chance of getting a trophy? Personally, I’d say no, and I’d want to see what, say, a 97% handicap percentage does. Here are those numbers: 25.9%, 23.5%, 21.0%, 18.7%, 16.1%, 15.3%, 13.6%, 13.0%, 11.6%, 10.1%, 41.8%. Is this even better? You decide.

This is the essence of tuning up a league.

[bowl1820]

I had to go back through my league sheets to find the data for my company leagues.

Okay summarizing that reply to my questions for the others.

1-What handicap was your league using then? and how many teams did it have?
Answer: 2006 season Handicap 90% of 210. Number of teams: 11 teams

2-What was your team average then? and what were the averages of the teams that won?

Answer: season ended with the teams having the following scratch averages (best-to-worst): 733, 708, 688, 665, 640, 631, 614, 598, 587, 561, 558. (I was on the 587 team. I ended with a 185 average, but my teammates obviously didn’t do nearly as well.)

3-When your ran the numbers for your league and it showed your company league was grossly unfair. What did it say would have been a fair handicap for your league?

This question isn't clearly answered.

this is what I feel is the main part that answers this question:

So a season in which a team both raised their average 20 pins and registered a trend of +8 would be a banner season for that team.

The breakout season for the poorest team^^

At this point, you’d try different handicap percentages to try to achieve a more competitive balance for this league. Going to 95% while continuing to assume a breakout season for the poorest team, we get these chances of getting a trophy: 33.6%, 27.8%, 23.8%, 19.9%, 16.4%, 14.5%, 12.3%, 10.5%, 9.6%, 7.2%, 34.6%. So this setting means that the best team bowling a typical season has the same chance of a trophy as the poorest team having a great season. That certainly sounds fairer to me!

The way I read it is, after making several assumptions about team performance. If the league went to 95% of 210 and the lower average teams bowled better. The lower average teams had more of chance to win.


So the main conclusion of this I see is that going from 90% to 95% only made a marginal change in where a team placed.

And what made a bigger difference was if the low average team increased their averages (Bowled better), they would have a better chance of winning.

Well DUH.

Is this the best percentage? That’s where subjectivity comes in. Some might say there’s still too strong a bias toward the better teams. And should a team having the season of its life still only have one-third of a chance of getting a trophy? Personally, I’d say no, and I’d want to see what, say, a 97% handicap percentage does. Here are those numbers: 25.9%, 23.5%, 21.0%, 18.7%, 16.1%, 15.3%, 13.6%, 13.0%, 11.6%, 10.1%, 41.8%. Is this even better? You decide.

Is going to 97% even better in this example? Yeah if you are the lowest average team. They have almost a 50% chance of winning, while everyone else is 25% or less!

I guess it's fair if you've always been in last place.

[bltuneup]

So the main conclusion of this I see is that going from 90% to 95% only made a marginal change in where a team placed.

And what made a bigger difference was if the low average team increased their averages (Bowled better), they would have a better chance of winning.

Well DUH.

Wow. I didn’t expect that kind of sarcasm in response. That’s clearly not my conclusion.

I’m not comparing how a team’s chances would be affected by doing better. I’m comparing how changes to a league’s handicap percentage affect the competitive balance of a specific league.

Conclusion 1: The best team in that league would expect a trophy more seasons than not, and over time could expect to get 35 trophies for every one trophy the worst team could expect. This does not appear to me to be fair. In fact, as I stated, I believe it’s “grossly unfair.” I personally would not have joined this league had I seen these numbers going in.

Conclusion 2: If the poorest team in the league did great, they would still have only an 18% chance of getting a trophy. Moreover, that team, bowling the best season it’s ever likely to have, still would only have one-third the chance of getting a trophy that the highest-average team had when bowling an unremarkable season. This reinforces conclusion 1’s assessment of the league’s inherent unfairness.

Conclusion 3: Raising the handicap percentage in this league will produce a more competitive balance that will give poorer teams a realistic shot at being rewarded for good performances, while still giving the higher-average teams an appropriate and fairly significant advantage. The fair amount to raise the percentage can be determined subjectively based on the analysis results I listed. I thought I was pretty clear in making the case that raising this league’s handicap percentage to 95% or 97% would both have been reasonable choices to make the league significantly fairer.

I should also add that this league disbanded a few years ago because the teams that never won a thing (including my own) eventually saw the futility of hoping for success and dropped out. My company could no longer convince enough teams to join, so the league disintegrated.

[bowl1820]

Wow. I didn’t expect that kind of sarcasm in response. That’s clearly not my conclusion.

I’m not comparing how a team’s chances would be affected by doing better. I’m comparing how changes to a league’s handicap percentage affect the competitive balance of a specific league.

I didn't think it was that sarcastic other than the Duh.

While You may not have intended to compare how a team’s chances would be affected by doing better, that's what happened.

First time you ran the numbers with no changes to handicap or trend. low was 0.7% chance of winning

Second time you ran the numbers, you had the low ave. team raise averages. This showed a change to 17.8%. in the lower average team chances of winning. So bowling better increased their chances of winning by 17.1 just by bowling a few extra pins all season.

Third time you ran the numbers, you added in a new handicap of 95%. This increased the chances of winning to 34.6% by 17.5.

I admit I was wrong earlier in saying that bowling better did more good than raising handicap in this instance (In this big mass of text it's hard to keep track of which line your looking at). But given if they didn't bowl better, that 34% would have dropped. The other teams percentages I assume would have raised up when the numbers ran. So the affect is still basically the same.

So the conclusion that shows is that while changing handicap flattens the chance of winning percentages some for the whole league. The thing that tips the scales for a teams chances of winning was a team increasing their average.

Conclusion 1: The best team in that league would expect a trophy more seasons than not, and over time could expect to get 35 trophies for every one trophy the worst team could expect. This does not appear to me to be fair. In fact, as I stated, I believe it’s “grossly unfair.” I personally would not have joined this league had I seen these numbers going in.

Now this is the area of being subjective as to what is fair.

You are saying here, that better teams have more of a chance to win/get trophies than lower average teams. .

I wouldn't say that is inherently unfair, I think most people would assume that better teams would win more.

Conclusion 2: If the poorest team in the league did great, they would still have only an 18% chance of getting a trophy. Moreover, that team, bowling the best season it’s ever likely to have, still would only have one-third the chance of getting a trophy that the highest-average team had when bowling an unremarkable season. This reinforces conclusion 1’s assessment of the league’s inherent unfairness.

Just because a low average team has it best season, doesn't necessarily mean that they are on par with the higher ave. bowlers. And that they should have higher chance of winning (ie: the 41.8% example) than the high average team has (ie: the 25.9% example)

Conclusion 3: Raising the handicap percentage in this league will produce a more competitive balance that will give poorer teams a realistic shot at being rewarded for good performances, while still giving the higher-average teams an appropriate and fairly significant advantage. The fair amount to raise the percentage can be determined subjectively based on the analysis results I listed. I thought I was pretty clear in making the case that raising this league’s handicap percentage to 95% or 97% would both have been reasonable choices to make the league significantly fairer.

As for Raising a league’s handicap percentage to make the league significantly fairer. That's what most all leagues have done over the years. just for that reason. The percentages years ago were only 70-80%, now we have 90-100% league handicaps.

I should also add that this league disbanded a few years ago because the teams that never won a thing (including my own) eventually saw the futility of hoping for success and dropped out. My company could no longer convince enough teams to join, so the league disintegrated.