CutDownTheNet
Well-known member
- Joined
- Oct 4, 2018
- Messages
- 2,547
- Reaction score
- 730
- Points
- 113
Possible New Seasonal Free-Throw Percentage Top-Ten Records
Summary:
(a) This year, there may be 2-4 free-throw percentage records added to the record books by players Pitts, Scalia, Powell and/or Brunson.
(b) Background on the why and how of that, and current record holders for comparison.
(c) A statistical Catch-22 that may (wrongfully) prevent these players from taking their rightful place in the record books.
(d) The fix for (c) and some philosophical aspects.
(a)
As of December 16 and heading into game 11 with Lehigh, both Destiny Pitts and Sara Scalia are shooting .909 from the free-throw line. On the odd chance that they might be able to maintain that statistic throughout the remainder of the season, they should (but see part (c)) displace both Rachel Banham (who owns the top 3 spots at .895, .874, .859) and Lindsey Whalen (4th spot at .847) as free-throw percentage queens. Destiny herself currently owns the 5th spot (.838), Lindsey the 7th spot (.832), and Rachel the 9th spot (.817) - which should all get bumped down the relevant number of slots.
A similar statement could be made about the free-throw shooting of Jasmine Powell (currently .838) and Jasmine Brunson (currently .833), both of whom might also join the top-ten seasonal free-throw percentage list, but are on the bubble at this point as to whether the might make, or fall short of, the Top Ten.
In other words, there’s a chance that as many as four of the 2019-20 Gopher players could join (or get an extra entry in, in the case of Pitts) the list of top-ten free-throw shooters by percentage.
(b)
The possibility that four current players could shoot free throws as well as the top-ten record holders, just boggles the mind. Of course, don’t count your free throws before they hatch. But it gives these four players a goal to aim at.
I believe that the “why” of this phenomenon boils down to the availability of the NOAH Basketball feedback system https://www.noahbasketball.com in our practice court. This helps the players fine-tune their shot, whether it be two-pointer, three-pointer or free throw. Many thanks go to the Fast Break Club for raising $10K to purchase this technology. (Commercial break: Consider joining the Fast Break Club - it only costs $50, but you’re free to give more if you like).
(c preview)
Unfortunately, however, there’s one gotcha that may prevent the proper updating of the record book, thus preventing all or part of (a) from happening.
Making that top-ten list (currently) has the overly stringent (and statistically unsound) prerequisite of shooting at least 100 free throw attempts on the season. Because we’re shooting a lot of free throws this year yet we’re spreading them around somewhat uniformly (i.e., it’s not like a past year in which Rachel Banham or Kenisha Bell took the lion’s share of free throws), it currently looks like only Jasmine Powell and Destiny Pitts have hopes of breaking that 100-attempt mark by end of season. Just not enough attempts to go around (without a lot more games like UC Davis) such as to give all four the minimal 100 attempts.
(b again)
We’re about one-third of the way through the season, so let’s guess estimated total seasonal free-throw attempts by tripling current attempts in the following table of estimated seasonal free-throw percentage for our current top-6 free-throw shooters ...
Estimated (projected) end-of-season free-throw percentage and estimated attempts
Name Est FTA Est FT% Adjusted FT%
Pitts 99 .909 TBD
Scalia 33 .909 TBD
Powell 111 .838 TBD
Brunson 54 .833 TBD
Hubbard 75 .760 TBD
T. Bello 135 .756 TBD
Taiye will have the mandated 100 attempts, but would not make the top-ten list unless she improves her current .756 to at least .830. Note however, that her free-throw percentage has improved from prior years - something that will help us win games.
Hubbard would need to improve her current .760 to at least .830 to make the record list, but at her current pace of free-throw attempts, might not make the 100 attempts prerequisite.
Brunson is borderline for having about a 10th place FT%, but at her current pace would not meet the 100-rebound prerequisite, so might get ripped off.
Powell currently has the best chance of making the top-ten list since it looks like she’s on pace for over 100 free throws at a qualifying .838 percentage that matches Destiny’s current 5th-place record. If she can match that .838 percentage in the rest of the season, she will be a new face in the record books.
It looks like Scalia is likely to get ripped off from becoming the new 1st place free-throw shooter by percentage (knocking Rachel and everybody else down a slot) since it looks like she will only get a third of the required 100 FTAs, in spite a new-high estimated .909 free-throw percentage (obtained by simply projecting her current percentage into the remainder of the season). This would be a shame. Admittedly there is less data to work with, but she might end up 14 points ahead of Rachel’s current 1st-place record. Is it fair to ban her from the record books just due to a smaller data sample? (More on that, below.)
The most controversial estimated result is for Pitts. She is estimated to tie Scalia for a new all-time high free-throw percentage of .909 - which (if it pans out that way) is clearly deserving of being the new first place free-throw shooter by percentage. Yet she might not get that honor merely because she ends up one free-throw attempt short of the 100-attempt requirement. If so, she would be getting severely ripped off - even more so than Scalia.
(c again)
The problem is the hard-cutoff at a minimum of 100 free-throw attempts. Is 99 attempts not close enough? So what the record-keepers are saying is that Destiny might end up with the highest free-throw percentage ever (albeit possibly tying with Sara), and yet not allowed to take that top spot in the free-throw percentage recordbook, simply because she was a silly FTA short of the (extremely arbitrary) minimal attempts required.
The 100-attempt requirement is not fair, and is set too high, and is statistically unsound as well (see part (d)). Probably a much smaller minimum (TBD but maybe about 25) attempts should be required, but other than that, small sample size should simply be factored into the ranking by means of using sample-size-discounted Adjusted FT% (the right-most column above) for ranking purposes (see part (d)).
By using such a (more statistically sound) approach, Destiny’s one-shy-of-100 99 projected attempts would not discount her estimated .909 raw percentage very much, and she almost certainly would become the new first-place free-throw shooter by percentage - which is as it should be if indeed she shoots free throws that well.
For Sara Scalia, her Adjusted FT% would be quite a bit lower to properly account for the smaller data sample, and she may or may not make the Top Ten list due to the discounting of Adjusted FT%, but that result would be statistically legitimate by ranking on Adjusted FT% instead of Raw FT%. At least she would have a chance of being rewarded for her strong effort by making the record book.
The record-book page could list both, such as Raw FT% (Adjusted FT%) - but sort on Adjusted FT% for the actual ranking (with suitable explanatory footnote). The Adjusted FT% would mathematically represent a lower bound on what FT% would be if the player had shot a pseudo-infinite number of free throws (like, say, thousands) on the season. We would retro-actively compute Adjusted FT% on existing record holders for the new sort.
(d)
Note that this statistical “issue” is only a problem on percentage statistics such as shooting percentages. It is not a problem with pure-number stats, since those represent the facts, and that’s that. It’s probably not a problem with X per game or Y per minute ratios either (with reasonable caveats). But shooting percentages intend to portray how good a shooter the player actually is. Philosophically, the FT% metric is only an estimate of what percentage a player would shoot free throws if they shot even more free throws on the season. Even the FT% statistic on a player that did shoot 100 free throws is still just an estimate of their free-throw shooting ability. A sample size of thousands of shots is really needed to be more accurate. (But of course, that is not possible.) Thus, the traditional (lazy) approach is just to set a fixed cutoff of minimum FTAs. Certainly you don’t name a player who shot 1-1 on the season the free-throw champion at 1.000. But where do you set the cutoff? At 5? At 10? At 20? At 50? At 100? At 200? At 500? The answer is that the hard-cutoff approach is simply too simplistic, and could have the effect of denying Destiny Pitts her well-deserved second appearance on the top-ten free-throw shooters by percentage.
A sample size of (as projected for Destiny) 99 shots is only microscopically less accurate than a sample size of 100. That’s the fundamental reason why a hard cutoff at 100 required shots is just “not right.”
And accuracy goes up only proportional to the square root of the sample size. So a sample size of 100 shots has an error margin that is only 1.414 times smaller than a sample size of 50 shots. Thus, discounting FT% to Adjusted FT% (to account for sample size) is both much more fair as well as much more statistically legitimate.
There are separate stats on number of free throws taken and number made in a season. Kenisha Bell is the champion in those departments. But when you divide free-throw makes by attempts, Bell is no longer in the top ten. The free throws taken and made are simple, pure, raw statistics, and so don’t have the statistical conundrum that FT% does. However, the way the 100-minimum shots policy now works, FT% is a contest limited to only the leaders in the free-throws taken category.
I’m convinced that the Discount FT% method is doable as well as much more fair. The statistical algorithm to implement it needs some more development and testing, but that could be accomplished before the end of the season.
At end of season I’ll be checking up on this, and (if needs be to give Destiny and/or Sara or others their due recognition) petitioning the Athletics Department and Whalen to fix the statistically broken FT% records page.
Summary:
(a) This year, there may be 2-4 free-throw percentage records added to the record books by players Pitts, Scalia, Powell and/or Brunson.
(b) Background on the why and how of that, and current record holders for comparison.
(c) A statistical Catch-22 that may (wrongfully) prevent these players from taking their rightful place in the record books.
(d) The fix for (c) and some philosophical aspects.
(a)
As of December 16 and heading into game 11 with Lehigh, both Destiny Pitts and Sara Scalia are shooting .909 from the free-throw line. On the odd chance that they might be able to maintain that statistic throughout the remainder of the season, they should (but see part (c)) displace both Rachel Banham (who owns the top 3 spots at .895, .874, .859) and Lindsey Whalen (4th spot at .847) as free-throw percentage queens. Destiny herself currently owns the 5th spot (.838), Lindsey the 7th spot (.832), and Rachel the 9th spot (.817) - which should all get bumped down the relevant number of slots.
A similar statement could be made about the free-throw shooting of Jasmine Powell (currently .838) and Jasmine Brunson (currently .833), both of whom might also join the top-ten seasonal free-throw percentage list, but are on the bubble at this point as to whether the might make, or fall short of, the Top Ten.
In other words, there’s a chance that as many as four of the 2019-20 Gopher players could join (or get an extra entry in, in the case of Pitts) the list of top-ten free-throw shooters by percentage.
(b)
The possibility that four current players could shoot free throws as well as the top-ten record holders, just boggles the mind. Of course, don’t count your free throws before they hatch. But it gives these four players a goal to aim at.
I believe that the “why” of this phenomenon boils down to the availability of the NOAH Basketball feedback system https://www.noahbasketball.com in our practice court. This helps the players fine-tune their shot, whether it be two-pointer, three-pointer or free throw. Many thanks go to the Fast Break Club for raising $10K to purchase this technology. (Commercial break: Consider joining the Fast Break Club - it only costs $50, but you’re free to give more if you like).
(c preview)
Unfortunately, however, there’s one gotcha that may prevent the proper updating of the record book, thus preventing all or part of (a) from happening.
Making that top-ten list (currently) has the overly stringent (and statistically unsound) prerequisite of shooting at least 100 free throw attempts on the season. Because we’re shooting a lot of free throws this year yet we’re spreading them around somewhat uniformly (i.e., it’s not like a past year in which Rachel Banham or Kenisha Bell took the lion’s share of free throws), it currently looks like only Jasmine Powell and Destiny Pitts have hopes of breaking that 100-attempt mark by end of season. Just not enough attempts to go around (without a lot more games like UC Davis) such as to give all four the minimal 100 attempts.
(b again)
We’re about one-third of the way through the season, so let’s guess estimated total seasonal free-throw attempts by tripling current attempts in the following table of estimated seasonal free-throw percentage for our current top-6 free-throw shooters ...
Estimated (projected) end-of-season free-throw percentage and estimated attempts
Name Est FTA Est FT% Adjusted FT%
Pitts 99 .909 TBD
Scalia 33 .909 TBD
Powell 111 .838 TBD
Brunson 54 .833 TBD
Hubbard 75 .760 TBD
T. Bello 135 .756 TBD
Taiye will have the mandated 100 attempts, but would not make the top-ten list unless she improves her current .756 to at least .830. Note however, that her free-throw percentage has improved from prior years - something that will help us win games.
Hubbard would need to improve her current .760 to at least .830 to make the record list, but at her current pace of free-throw attempts, might not make the 100 attempts prerequisite.
Brunson is borderline for having about a 10th place FT%, but at her current pace would not meet the 100-rebound prerequisite, so might get ripped off.
Powell currently has the best chance of making the top-ten list since it looks like she’s on pace for over 100 free throws at a qualifying .838 percentage that matches Destiny’s current 5th-place record. If she can match that .838 percentage in the rest of the season, she will be a new face in the record books.
It looks like Scalia is likely to get ripped off from becoming the new 1st place free-throw shooter by percentage (knocking Rachel and everybody else down a slot) since it looks like she will only get a third of the required 100 FTAs, in spite a new-high estimated .909 free-throw percentage (obtained by simply projecting her current percentage into the remainder of the season). This would be a shame. Admittedly there is less data to work with, but she might end up 14 points ahead of Rachel’s current 1st-place record. Is it fair to ban her from the record books just due to a smaller data sample? (More on that, below.)
The most controversial estimated result is for Pitts. She is estimated to tie Scalia for a new all-time high free-throw percentage of .909 - which (if it pans out that way) is clearly deserving of being the new first place free-throw shooter by percentage. Yet she might not get that honor merely because she ends up one free-throw attempt short of the 100-attempt requirement. If so, she would be getting severely ripped off - even more so than Scalia.
(c again)
The problem is the hard-cutoff at a minimum of 100 free-throw attempts. Is 99 attempts not close enough? So what the record-keepers are saying is that Destiny might end up with the highest free-throw percentage ever (albeit possibly tying with Sara), and yet not allowed to take that top spot in the free-throw percentage recordbook, simply because she was a silly FTA short of the (extremely arbitrary) minimal attempts required.
The 100-attempt requirement is not fair, and is set too high, and is statistically unsound as well (see part (d)). Probably a much smaller minimum (TBD but maybe about 25) attempts should be required, but other than that, small sample size should simply be factored into the ranking by means of using sample-size-discounted Adjusted FT% (the right-most column above) for ranking purposes (see part (d)).
By using such a (more statistically sound) approach, Destiny’s one-shy-of-100 99 projected attempts would not discount her estimated .909 raw percentage very much, and she almost certainly would become the new first-place free-throw shooter by percentage - which is as it should be if indeed she shoots free throws that well.
For Sara Scalia, her Adjusted FT% would be quite a bit lower to properly account for the smaller data sample, and she may or may not make the Top Ten list due to the discounting of Adjusted FT%, but that result would be statistically legitimate by ranking on Adjusted FT% instead of Raw FT%. At least she would have a chance of being rewarded for her strong effort by making the record book.
The record-book page could list both, such as Raw FT% (Adjusted FT%) - but sort on Adjusted FT% for the actual ranking (with suitable explanatory footnote). The Adjusted FT% would mathematically represent a lower bound on what FT% would be if the player had shot a pseudo-infinite number of free throws (like, say, thousands) on the season. We would retro-actively compute Adjusted FT% on existing record holders for the new sort.
(d)
Note that this statistical “issue” is only a problem on percentage statistics such as shooting percentages. It is not a problem with pure-number stats, since those represent the facts, and that’s that. It’s probably not a problem with X per game or Y per minute ratios either (with reasonable caveats). But shooting percentages intend to portray how good a shooter the player actually is. Philosophically, the FT% metric is only an estimate of what percentage a player would shoot free throws if they shot even more free throws on the season. Even the FT% statistic on a player that did shoot 100 free throws is still just an estimate of their free-throw shooting ability. A sample size of thousands of shots is really needed to be more accurate. (But of course, that is not possible.) Thus, the traditional (lazy) approach is just to set a fixed cutoff of minimum FTAs. Certainly you don’t name a player who shot 1-1 on the season the free-throw champion at 1.000. But where do you set the cutoff? At 5? At 10? At 20? At 50? At 100? At 200? At 500? The answer is that the hard-cutoff approach is simply too simplistic, and could have the effect of denying Destiny Pitts her well-deserved second appearance on the top-ten free-throw shooters by percentage.
A sample size of (as projected for Destiny) 99 shots is only microscopically less accurate than a sample size of 100. That’s the fundamental reason why a hard cutoff at 100 required shots is just “not right.”
And accuracy goes up only proportional to the square root of the sample size. So a sample size of 100 shots has an error margin that is only 1.414 times smaller than a sample size of 50 shots. Thus, discounting FT% to Adjusted FT% (to account for sample size) is both much more fair as well as much more statistically legitimate.
There are separate stats on number of free throws taken and number made in a season. Kenisha Bell is the champion in those departments. But when you divide free-throw makes by attempts, Bell is no longer in the top ten. The free throws taken and made are simple, pure, raw statistics, and so don’t have the statistical conundrum that FT% does. However, the way the 100-minimum shots policy now works, FT% is a contest limited to only the leaders in the free-throws taken category.
I’m convinced that the Discount FT% method is doable as well as much more fair. The statistical algorithm to implement it needs some more development and testing, but that could be accomplished before the end of the season.
At end of season I’ll be checking up on this, and (if needs be to give Destiny and/or Sara or others their due recognition) petitioning the Athletics Department and Whalen to fix the statistically broken FT% records page.
Last edited: