BBKL GM Vote (Yes or No): Adding PPTOI from 2014/2015

[Nick]

To be fair shook, I won more weeks* (not categories) then anyone last year, with a team geared totally towards scoring... dmen that score, centres that score & wingers that score.

you have gone with that build since the beginning of bbkl. I am not trying to insult you but you have yet to win a playoff round..

yep i have... and not that this should need to be said, but timing of injuries is a rather big deal. Never started the playoffs with my starting goalie until this year; without Crosby and co this year... troll attempt perhaps, but the team I has the last two seasons could easily have won with a bounce a different way.

You did a 180 change this offseason and now have a periph based team, rather then an offensive one. I think you giving up on a build that all others had currently done is a sign that yes you can be competitive but you can't win. If it was strictly injuries, now I realize cap accounts for the trades just not the pieces targeted, why would you change builds?

I'll return to a scoring build in a year or two, had to make a change in build in order to make any trades..otherwise i would have just stood pat with what I had. I don't think I've lost the core of a scoring team by any measure. I switched to non-scoring wingers, which should make my total season points (our standings) look better, but I doubt it helps me win those hard weeks. G/A/Pts/Stpts/SOG are still the biggest group of commonly corelated categories.

for the record, Shiv, Shep and Steve had the same record as you. purely W/L record.

Missed what I said, with schedule removed as a factor, I would have won more weeks against everyone else last year. Not won more categories, but won more weeks with the confounding effecot of scheduled removed. I didn't get to play Shiv or Shep, but did beat Steve fwiw.

[Robin Hood]

I had a winning record against ya Nick. 107-91-28.

My only losing records were against KyleBOS (94-108-22) and Mike (88-115-21), both of whom have teams built similar to mine and Sheps.

Against Bergey I was 122-82-20 but in the weeks we actually played each other I lost 13-17-2 lol. That first round loss still hurts.

[Nick]

yes, you won more categories against me over the year, not on weekly based outcomes. I also beat Kyle and Mike last year Shiv, and would have more weeks than I lost. Clearly showing the error in our standings system.

[shooker]

I knew what you meant as in just weeks as a 1 win per. I thought you meant you won more weeks as in record over the season. There were 4 teams including yourself at 9-1. I get what you mean now.

Injuries 5 years in a row? actually I remember that epic matchup with Kareem and Yourself a while back. Both would have beat any other team in the league that week lol. I am just chirping you now.

[Nick]

I knew what you meant as in just weeks as a 1 win per. I thought you meant you won more weeks as in record over the season. There were 4 teams including yourself at 9-1. I get what you mean now.

Injuries 5 years in a row? actually I remember that epic matchup with Kareem and Yourself a while back. Both would have beat any other team in the league that week lol. I am just chirping you now.

Would not claim first 2 years, I wasn't in it to win it, had a prospect based team with a build plan that worked pretty well IMO. last 2 years for sure, and I would have needed to be lucky 3 years ago, but not having a starting goalie due to injuries for years 1-4 is simply unreal timing. starting a week down 0-4 against other quality opponents is not a recipe for success.


FTR -> Emery as a flyer (hip disaster), Hiller (vertigo) & Howard (Knee?) all with injuries during the first round of our playoffs.

[Robin Hood]

yes, you won more categories against me over the year, not on weekly based outcomes.

If you had the edge in a weekly H2H setting against me, you would be winning more categories...not sure where your logic is going lol.

[shooker]

your goaltending might actually jump to elite bbkl wise this year. The acquisitions Detroit made + great team defense + less travel. They are going to be a hard team to beat this season.

anyways back on topic you two.

[Nick]

yes, you won more categories against me over the year, not on weekly based outcomes.

If you had the edge in a weekly H2H setting against me, you would be winning more categories...not sure where your logic is going lol.

Somewhere I already posted the outcomes if we had a weekly based record outcome. You are making the mistake of magnitude of win, which is a flaw in our current system.

There were two weeks you dominated me last year, 14-1-1 and 11-3-2 or something. As well as two weeks of tight (9-7 sorta thing). I never earned 11 categories off of you, but won more weeks by the smaller amounts, never overcoming the -10 and -8 of those two big weeks.


Not running the # again as it's all moot, no one cares to change our standing system (whether by choice or not understanding, i'm not sure) but:

Example:
PIT VS TBL
(1)11-3-2
(2)9-7
(3)9-6-1
(4)14-1-1
(5)7-8-1
(6)7-9
(7)6-8-2
(8)7-9
(9)7-8
(10)7-7-2
--------------------
weekly thats a 4-5-1 record, but categorically thats an 84-66-9

[Robin Hood]

yes, you won more categories against me over the year, not on weekly based outcomes.

If you had the edge in a weekly H2H setting against me, you would be winning more categories...not sure where your logic is going lol.

Somewhere I already posted the outcomes if we had a weekly based record outcome. You are making the mistake of magnitude of win, which is a flaw in our current system.

There were two weeks you dominated me last year, 14-1-1 and 11-3-2 or something. As well as two weeks of tight (9-7 sorta thing). I never earned 11 categories off of you, but won more weeks by the smaller amounts, never overcoming the -10 and -8 of those two big weeks.

But this implies that in the long run I win more categories against you than vice versa. The anomalies aren't outliers. They're just how the distribution happened.

[Nick]

No it does not.

[shooker]

122-82-20 over 14 weeks is only an average of a 9-6-1. That is actually a lot closer then the score looks. but enough dick measuring, lets get back to the topic at hand.

[Nick]

But this implies that in the long run I win more categories against you than vice versa. The anomalies aren't outliers. They're just how the distribution happened.

It's measure magnitiude of the win, not the # of wins. It's saying winning if the leafs beat the bruins by 6 goals in one game, and for the next 5 that the leafs beat the bruins by 1, that the record is 6-5.

[Robin Hood]

But this implies that in the long run I win more categories against you than vice versa. The anomalies aren't outliers. They're just how the distribution happened.

It's measure magnitiude of the win, not the # of wins. It's saying winning if the leafs beat the bruins by 6 goals in one game, and for the next 5 that the leafs beat the bruins by 1, that the record is 6-5.

It's not at all the same thing lol.

In fact, it's the opposite. It's the equivalent of you flipping a coin 1000 times then segmenting the results into groups of 10. Then saying that Heads won more groups of 10 than Tails even though the total distribution was 50%.

[Nick]

But this implies that in the long run I win more categories against you than vice versa.

It's a factor of what the wins look like. because you win 5 categories that i do not compete in does not mean you win more weeks, when out of 12 skater categories i dominate 7 of them (although you were elite as well). So in a week that you stole G & GWG + your 5 and a sweep of goalie stats = big win. But the much more common outcome, was my 7 dominant categories + 1 other (whether goalie, +/- or w/e) resulted in a win for me.

We run a weekly comparison league, the measure SHOULD be who won the week, not by how much they won a week by comapred acorss the league/season. It's actually a large flaw that we don't have an interest in fixing

[shooker]

also we need magnitude of wins in this league, with 21 weeks there would be too many ties. Real life the top 3 or 4 teams in each conference (there are exceptions: LA) tend to score more goals, allow less and then win by wider margins on a year round average. I understand the logic behind 1 win per week but this breaks ties and gives higher seats to those who dominated on a week to week basis. I think that should be rewarded. With the schedule being very balanced between divisions it should properly depict that.

False. Already showed that there would be less extreme seperation in the league but more accurate season pictures a race in the 6-10 spot in the last 3 seasons. sorry, hit edit not quote.

[Nick]

Shiv, I am 110% sure, that we measure magnitiude of wins, not # of wins. You're making a statistical error and I've already posted on this to the point of nasuem in the CC. A team that won every single week with a score of 9-7 for a perfect 21-0 weekly record wouldn't even win their division in our standing system. You winning 17 weeks by 14-1-1, and losing 4 weeks 7-9 should not place you higher in the standings. It does.

[Robin Hood]

But this implies that in the long run I win more categories against you than vice versa.

It's a factor of what the wins look like. because you win 5 categories that i do not compete in does not mean you win more weeks, when out of 12 skater categories i dominate 7 of them (although you were elite as well). So in a week that you stole G & GWG + your 5 and a sweep of goalie stats = big win. But the much more common outcome, was my 7 dominant categories + 1 other (whether goalie, +/- or w/e) resulted in a win for me.

We run a weekly comparison league, the measure SHOULD be who won the week, not by how much they won a week by comapred acorss the league/season. It's actually a large flaw that we don't have an interest in fixing

This is terrible terrible analysis. I'm shocked that you're a stat major and making this argument.

First - when you and I begin any game week, each category has a probability. For example, let's say you have an edge in PIMS, the odds of you winning the category that week are say 60%, mine are 35%, 5% chance of a tie (making these numbers up). Over 10 weeks - the split on the PIMS category should technically be 6-3-1 or 6-4-0 in your favour. The same goes for every category.

Second - the distribution of these categories does not "need" to happen as equally every week. That is, if you are a 60% favourite on the 12 skater categories, you do not need to win 7-8 categories every week. You do however at the end of the 10 weeks need to be close to the 60% mark of the 12 categories. So some weeks you may lose 8 out of 12. But you should win 7-9 in others to make up the difference.

Best way I can sum up the mistake you are doing is what I've already posted:

It's the equivalent of you flipping a coin 1000 times then segmenting the results into groups of 10. Then saying that Heads won more groups of 10 than Tails even though the total distribution was 50%.

The only logical argument you can make is that the sample size was smaller this season than the usual 25 weeks.

[shooker]

Its not perfect, but there is one maybe 2 teams a year that get hurt by the fact there is 16 wins in a week. Most of the time it just shows the difference in the teams, or worst case scenario, their schedules but that would be your fault facey

[Nick]

Sorry Shook, hit edit on your post, and not quote.


The seperation created by our standings system is not accurate. I've already done the work showing that even in JUST 10 weeks there are not too many ties, but over 21 weeks it only gets more and more accurate. It doesn't create false seperation or extreme gaps, and makes changes in the standings a result of # of weeks won, and not by how big of wins (or losses).


Saying a team can steal 2 or 4 points from a contender to make the week interesting is the dumbest defense. It's still a 14-12 point spread instead of a 2, Our scoring system makes it virtually impossible for a bad start to a seaosn to be overcome, in the very nature that a team that loses by 8+ categories is unlikely to win by 8+ categories, so the loss is not worth the same as the win. It's insanity.

[Robin Hood]

A team that won every single week with a score of 9-7 for a perfect 21-0 weekly record wouldn't even win their division in our standing system. You winning 17 weeks by 14-1-1, and losing 4 weeks 7-9 should not place you higher in the standings. It does.

This is a different argument. I agree with this when it comes to BBKL standings as a team that is above average in every category will likely outperform one specialized in 9-10.

However, when comparing 2 teams H2H, there is no external factor. It is comparing how many categories one won against the other. And in the long run, if Team A is better than Team B in 9-10 categories on average, then Team A should have an average of 9-10 wins over a season when the sample size is big enough. There is no issue of magnitude.