Line stats?

fredbeene · 11-23-2022, 12:10 AM

Are their stats in the game to asses which lines are best?

How would i know if my line 3 is actually more productive then line1?
ie tossing numbers out for period
line 3 is on ice 5 mins and scores 1 goal
line 2 on ice 5 mins and doesn't score
line 1 on ice 10 mins and scores 1 goad.

line 3 is actually more productive scoring.

Or do people not track it this way because there are so many players on and off the ice that lines aren't really tracked?

TobyInHR · 11-23-2022, 01:31 PM

First, I'm happy to say I am currently working on a Google Sheet that I'm almost ready to share with the community, which takes your game's CSV files and provides deeper analysis of your team using the stats we get from the game. You can go player-by-player, or line-by-line to see who is producing, who isn't, and why. I'd be happy to share that with you when it's ready. If you'd like to test it out before then, hit me up on Discord! You can find me in the OOTP FHM channel.

Second, I think your example shows a fundamental flaw in why we don't necessarily look at line production that way. You're dealing with a relatively small sample size, which means the results will not be easily replicated over the course of a whole season. This is why the go-to scale is to look at production per 60 minutes played, that way everyone is on the same level when we analyze them.

To illustrate, let's say your third line has 1 goal in 5 minutes of TOA, but they only took 10 shot attempts in those 5 minutes (Corsi For), and 3 of those shots got blocked before reaching the net, giving them a Fenwick For of 7, and a Fenwick For Shooting % of 14.2%.

We could project, then, that in 60 minutes of TOA, your third line would have 84 unblocked shot attempts, and score on 11 of them. But that projection is flawed -- a 14.2% shooting percentage is unsustainably high. In my most recent save, the top 5 3rd lines in the league by FFSH% were between 8.0% and 8.4%. We should assume your third line will regress to the mean over their next 60 minutes. So, we say they shot 14% in the first 5 minutes, and 8% in the next 55. Assuming they keep the same rate of shot attempts and Fenwick, that's 8% on 77 shots, or 6 goals, total of 7 goals/60.

Meanwhile, let's say your first line has 1 goal in 10 minutes, but they have 25 unblocked shot attempts (roughly 50% more shot attempts than your third line because skilled shooters play more minutes in general). First lines also average a much higher FFSH%, in my league between 9 and 10.5%. So, using the same projections, your first line underperformed in their 10 minutes, but will still likely regress to the mean: over 60 minutes they will have 150 unblocked shot attempts, and score a total of 13 goals.

Don't get caught thinking that line 3's 1 goal in 5 minutes means they're absolutely more productive than line 1 with 1 goal in 10 minutes. They are over that specific window, but there is much more to it than just extrapolating data from a small sample size.

Thankfully, my data sheet will do all this work for you! You'll be able to see instances where players are *actually* outperforming their line and should be moved up to see more ice time because their contributions are exponentially more impactful when given time to play.

renojedi · 11-23-2022, 02:59 PM

Great write up! I love the way OOTP had added advanced stats. Hope FHM can get there too...Corsi, Fenwick, and On-Ice Shooting Percentage at the minimum. Glad your spreadsheet can bridge the gap...looking forward to it!

Savoie45 · 11-24-2022, 01:39 PM

Quote:

Originally Posted by renojedi

Great write up! I love the way OOTP had added advanced stats. Hope FHM can get there too...Corsi, Fenwick, and On-Ice Shooting Percentage at the minimum. Glad your spreadsheet can bridge the gap...looking forward to it!

FHM has Corsi under advanced stats.

11-23-2022, 12:10 AM	#1
fredbeene All Star Starter Join Date: Aug 2016 Posts: 1,690 Infractions: 0/1 (1)	Line stats? Are their stats in the game to asses which lines are best? How would i know if my line 3 is actually more productive then line1? ie tossing numbers out for period line 3 is on ice 5 mins and scores 1 goal line 2 on ice 5 mins and doesn't score line 1 on ice 10 mins and scores 1 goad. line 3 is actually more productive scoring. Or do people not track it this way because there are so many players on and off the ice that lines aren't really tracked?

11-23-2022, 01:31 PM	#2
TobyInHR Minors (Rookie Ball) Join Date: Oct 2022 Posts: 30	First, I'm happy to say I am currently working on a Google Sheet that I'm almost ready to share with the community, which takes your game's CSV files and provides deeper analysis of your team using the stats we get from the game. You can go player-by-player, or line-by-line to see who is producing, who isn't, and why. I'd be happy to share that with you when it's ready. If you'd like to test it out before then, hit me up on Discord! You can find me in the OOTP FHM channel. Second, I think your example shows a fundamental flaw in why we don't necessarily look at line production that way. You're dealing with a relatively small sample size, which means the results will not be easily replicated over the course of a whole season. This is why the go-to scale is to look at production per 60 minutes played, that way everyone is on the same level when we analyze them. To illustrate, let's say your third line has 1 goal in 5 minutes of TOA, but they only took 10 shot attempts in those 5 minutes (Corsi For), and 3 of those shots got blocked before reaching the net, giving them a Fenwick For of 7, and a Fenwick For Shooting % of 14.2%. We could project, then, that in 60 minutes of TOA, your third line would have 84 unblocked shot attempts, and score on 11 of them. But that projection is flawed -- a 14.2% shooting percentage is unsustainably high. In my most recent save, the top 5 3rd lines in the league by FFSH% were between 8.0% and 8.4%. We should assume your third line will regress to the mean over their next 60 minutes. So, we say they shot 14% in the first 5 minutes, and 8% in the next 55. Assuming they keep the same rate of shot attempts and Fenwick, that's 8% on 77 shots, or 6 goals, total of 7 goals/60. Meanwhile, let's say your first line has 1 goal in 10 minutes, but they have 25 unblocked shot attempts (roughly 50% more shot attempts than your third line because skilled shooters play more minutes in general). First lines also average a much higher FFSH%, in my league between 9 and 10.5%. So, using the same projections, your first line underperformed in their 10 minutes, but will still likely regress to the mean: over 60 minutes they will have 150 unblocked shot attempts, and score a total of 13 goals. Don't get caught thinking that line 3's 1 goal in 5 minutes means they're absolutely more productive than line 1 with 1 goal in 10 minutes. They are over that specific window, but there is much more to it than just extrapolating data from a small sample size. Thankfully, my data sheet will do all this work for you! You'll be able to see instances where players are actually outperforming their line and should be moved up to see more ice time because their contributions are exponentially more impactful when given time to play.

11-23-2022, 02:59 PM	#3
renojedi All Star Reserve Join Date: May 2003 Location: Richmond, Virginia Posts: 502	Great write up! I love the way OOTP had added advanced stats. Hope FHM can get there too...Corsi, Fenwick, and On-Ice Shooting Percentage at the minimum. Glad your spreadsheet can bridge the gap...looking forward to it!