Login

Kami1931 · 05-18-2025, 05:07 AM

SB is incorrectly counting the blocker as the shooter. The blocker gets a SB for the blocked shot and a SQ value for the shot attempt. Big thanks to Luke for finding a dev post that references this issue. The article below is left as is but a reminder that my conclusion was incorrect.

Shots Blocked or Blocked Shots?

With the upgrade to FHM 10 there has been many new stats for us to sink our teeth into. Zone Starts, Shot Quality, and Team Shots. While working with Shot Quality I found numbers were not matching up to what I expected. After working on it for a bit I firmly believe the SB stat is not how many Shots the Player has blocked themself but instead is the amount of shots they've had blocked. Below is a bit of what I've found.

When I first started I noticed that SOG didn't match the shot quality total.

[Image: sl8kGUx.png]

That's fine right? I can just assume that the difference is Blocked Shots (BS) by the opposite team. Well, just to be safe I'll just check this game and if it works then I should be good to go. The Steelhawks have 22 SOG, probably be easier to do them first so let's get the BS from the Panthers.

[Image: ipLUAzJ.png]

Alright LAP has 16 BS, 16 + 22 = 40 Shot attempts for HAM. Closer but still 10 short. Missed shots (MS) aren't supposed to be included in shot quality right? I can look at players shot quality in game, if Shot Quality Total (SQT) is always higher than SOG+MS then I'll be almost certain that MS's have shot quality. At least one player is going to miss more shots than they have on goal and blocked. Let's grab the boxscore_skater_summary and check.

[Image: BkWTuaV.png]

Alright there's only.... 17542 Rows. Fun. And it's the J stats, but that shouldn't matter. I'll just bring the stats over I need SOG, MS, SQT, then use the IF function. It will say yes when SQT >= MS then I can use the CountIf function to count both the YES's and No's. Wow, okay there isn't one time out of the 17542 rows that a player had more missed shots than SQT. Missed shots probably have shot quality. This is unexpected I better reach out and see if someone knows something. Luke responded that they believed MS shouldn't be included based on the model that FHM 10 used for determining shot quality. He also mentioned something was wrong with blocked shots not matching up. Great, more problems. Cake also said it was unlikely as her player had lots of high quality chances (sq4) but a terrible shooting percentage. I'll dig and post a reply (with terrible team shot quality math in hindsight!).

https://index.simulationhockey.com/shl/8...?season=82

Alexa Johansen had 5 SOG, 1 BS, 2MS (from YouTube), and 8 SQT. Wait, 5+1+2 = 8. Maybe it's by chance that she had exactly one shot blocked and one blocked shot? That's weird, I'll check that out later.

Now being fairly confident that missed shots are included in SQT. Let's use the sim channel and grab the missed shots by HAM and see if it solves my problem.

[Image: s4amBFE.png]

1 + 1 + 3 = 5.... I'm 5 Shot Attempts short. Am I dumb? Yes, is my math bad? After checking a couple times it's okay this time. I'm at a loss, maybe there's something wrong with FHM and missed shots. Better ask people that are more experienced than me, maybe they know something. Luke confirmed something was up with blocked shots in FHM10. Great, it might just be a problem with FHM 10 and there's nothing to do. But what about Alexa? Would changing the BS stat from blocked shots to shots that were blocked fix my problem? Let's try again with HAM's BS instead of LAP's

Back to HAM. 22 SOG, 23 BS, 5 MS for a total of 50! There it is! That's the number I wanted, but how sure am I that this is true all the time? Not very and BS always means the amount of shots the player has blocked. Gotta go big (data).

I need the boxscore_summary. It has GameIds, TeamIds, Shots Home and Away, and Shot Quality. All that's left are the Blocked Shots and Missed shots from boxscore_skater_summary that I already have.

[Image: jcEzOfy.png]

Easy enough to get the base info so I'll just add a bunch of helper rows like SOG+MS, SOG+MS+BS, shot attempts so it's easier to understand. I'll check my idea first that SQT (Shot attempts) = SOG+MS+BS for every game (all 584 of them). The row will say Yes when they are equal and wow. Every game, both the home and away teams SQT = SOG + (Their Players) BS, + MS. BS is very likely to be the shots the player has had blocked. But I need another way to prove it, probably through player data.

IF BS is the shots the player has blocked themself, There should exist a player that at some point blocked more shots than attempted in a game. A player could block 5 shots and only attempt 2. Seems very possible. I just need to test BS > SQT for every player in every game. Sounds worse than what it is.

[Image: eOpRdEL.png]

Here we have each player in each game's SOG,MS,BS,SQ0-4,SQT brought over from the boxscore_skater_summary. Now to check if there's ever been a player with more BS than SQT, and there's not one in 17542 times players played in games. What are the odds of that? I'll use an easy binomial function with a few values until it says it's likely possible. There's 17.5% chance of 0 players having more blocked shots than SQT if the odds of having a player having more BS than SQT in a game is one in ten-thousand. In other words, in 667 333 Games you can expect one player to have more BS than SQT. I think that should happen a little more often than that. Therefore, SB is very likely to be shots the player has had blocked.

After undertaking this I firmly believe SB is the shots attempted by the player that were blocked. It fits all the data and (from my limited experience) matches the shot quality overlay in FHM 10.

What does this mean for SB? Well, It's no longer a measure of any defensive contribution. However, it could be used to better determine players that aren't able to get shots through as well as tactics that decrease SB for the team.

A special thanks to @luke and @cake307 for answering my questions on Discord and giving me the right leads that eventually led me to this conclusion.

P.S

What about SQ0>(BS+MS)? It tests for when players have had more sq0 than BS + MS. BS are known to derate a shots quality by one, my assumption is that MS do as well. So how do you have SQ0 more than what causes it? I believe that it's the shootout, but more on that later.

Edit: I used 15 skaters (One team) when calculating how often this would happen. Two teams, 30 skaters actually play. The correct math is now shown, 10,000/30 = 333.33.

Ronniewalker · 05-18-2025, 06:16 AM

Very interesting! So SB has changed from a defensive stat to an offensive one? And nothing in FHM 10 support information actually discusses this?

Also, what does this mean for fantasy scoring?

Kami1931 · 05-18-2025, 06:37 AM

05-18-2025, 06:16 AMRonniewalker Wrote: Very interesting! So SB has changed from a defensive stat to an offensive one? And nothing in FHM 10 support information actually discusses this?

Also, what does this mean for fantasy scoring?

I believe so yes. The only thing I found was in the manual:

Shot Blocking (Defensive)
The willingness and skill level of the player when trying to block shots. Defencemen should normally tend to get higher ratings than forwards, which balances out the extra defensive points the forwards get from the Faceoff attribute. If the player's league records shot-blocking stats, these would be very useful in setting this attribute.

Which suggests this may be an accident or oversight by the developers and not intended. It is also unclear how long this has been in game. This may have been present in fhm 8 but without shot quality there was no way to find it. I will (time pending) attempt to find out if it's possible at a later date.

For fantasy I think using it for defenders past this year would be a mistake. Which would require a complete rebalanced of the point system. I do not envy the person who does it.

***luke*** · 05-18-2025, 08:53 AM

When I first found this when I was putting the FHM10 info into the index. I thought it was just a fuck up on the devs. That some numbers slightly mismatched based on what the actual in game showed. But with Kami’s great research, it looks like they really fucked up.

https://index.simulationhockey.com/shl/8...?season=82

Taking a look at this game in the sim itself also shows a different story to just the exports.

So first we can take a look at all the blocked shots in the Shots Map

[Image: s2jBD6I.png]

Left hand side is SFP’s shots and right hand side is EDM’s shots.

We can take a look at Alexa Johansen like Kami has done

5 Shots on net, 2 missed shots, and 1 blocked shot

The 2 missed shots are accurately displayed on the map showing missed shots.
The 5 shots on net are also correct. He had 3 shots that were saves and 2 missed shots
The 1 blocked shot though?
On the map it only has 1 blocked shot event when Alexa Johansen shot the puck. Which was actually blocked by Oskar Scholz
There is also a situation where Alexa Johansen actually blocked a shot as well. In which Alexa blocked Fredrik Gronlund’s shot. Who in total shot the puck 3 times that were a blocked shot event

Edmonton has had 14 shots blocked
San Francisco has had 18 shots that were blocked.

Off of the same game there are a Team Stats Tab and a Shots tab in game we can look at to see more information

[Image: Lmmndhu.png]

For Edmonton’s Team in the Blocked Shots column they had 14 blocked shots
For San Francisco’s Team in Blocked shots column they had 18 blocked shots

But lets take a look at the Shot Attempts and Shots in total

Edmonton had 79 shot attempts and 52 actual shots

However if we take a look at the boxscore its a different story

[Image: JmOXKVw.png]

It is also interesting to note for this game as well Walt “Clyde” Frazier also actually blocked 2 shots according to the map

Second game was 2049-1-14
[Image: d20fS4H.png]

Third game is the same. It was on 2049-1-25 where he had 1 “shot blocked”
[Image: xkrTB8w.png]

Also in this game he did not record a shot blocked on the other side of the ice

Fourth game was 2049-3-1
[Image: JmBD6pl.png]

Again same situation where this was the only time Walt’s name was mentioned in the shots blocked category

I can confirm that Kami is correct in that I believe the FHM devs just fucked up. They put the shots that was blocked when the player shot as shots blocked. Instead of the actual blocked shot event that they do record but do not look like they do not record for whatever reason.

icewindoo · 05-18-2025, 08:59 AM

This is great sleuthing! Thanks for looking into this. I agree 100% that something is definitely wrong here.

But I'm starting to wonder if the part that's wrong is not the meaning of the 'SB' stat, but perhaps there's an error in how FHM is tallying up those total shot quality numbers.

I have two pieces of data to suggest that SB still really does mean "shots that were blocked by the defensive player".

First, I still see a fairly clear correlation between SB per minute and the player's Shotblocking attribute / rating:

[Image: G5XkEUB.png]

If the SB stat were really about shots taken by that player but which were blocked by the defender, then I would not expect to see any relationship between SB and the attribute. But that graph, IMHO, shows a clear connection.

Second, the team SB totals (which do match the sum of the player SB values) still indicate a real disconnect between SOG (here, 'S') and SB:

[Image: vKiArvF.png]

If SB was really about "shots taken that were blocked", then I'd expect to see a noticeable connection between SOG and SB: the more total shot attempts your team has, then both SB and SOG would rise. But that's not what the data shows. If anything, the inverse is true: the more SOG a team has, the fewer SB it has, which makes sense if SB were the defensive shot blocking measure.

So my theory is that the shot quality numbers are off somehow, probably somehow using defensive blocked shots in place of offensive shots blocked. (Good thing nobody did a media yet on shot quality!) But I can't see yet exactly where the error is.

Kami1931 · 05-18-2025, 10:47 AM

05-18-2025, 08:59 AMicewindoo Wrote: This is great sleuthing! Thanks for looking into this. I agree 100% that something is definitely wrong here.

But I'm starting to wonder if the part that's wrong is not the meaning of the 'SB' stat, but perhaps there's an error in how FHM is tallying up those total shot quality numbers.

I have two pieces of data to suggest that SB still really does mean "shots that were blocked by the defensive player".

First, I still see a fairly clear correlation between SB per minute and the player's Shotblocking attribute / rating:

If the SB stat were really about shots taken by that player but which were blocked by the defender, then I would not expect to see any relationship between SB and the attribute. But that graph, IMHO, shows a clear connection.

Second, the team SB totals (which do match the sum of the player SB values) still indicate a real disconnect between SOG (here, 'S') and SB:

If SB was really about "shots taken that were blocked", then I'd expect to see a noticeable connection between SOG and SB: the more total shot attempts your team has, then both SB and SOG would rise. But that's not what the data shows. If anything, the inverse is true: the more SOG a team has, the fewer SB it has, which makes sense if SB were the defensive shot blocking measure.

So my theory is that the shot quality numbers are off somehow, probably somehow using defensive blocked shots in place of offensive shots blocked. (Good thing nobody did a media yet on shot quality!) But I can't see yet exactly where the error is.

The first correlation, between SB/minute and shotblocking stat could be explained by two things.

One, the devs don't know this was a mistake and balanced the formula using the shotblocking attribute of the player taking the shot. Shotblocking could be an actual negative stat. I don't want to believe this because having another useless attribute just isn't fun but I have to admit that this is possible.

Two, higher TPE players could be getting more offensive zone starts. This leads to more shot attempts per minute played and more of their shots blocked. At least compared to the poor guy starting in the defensive zone all the time.

On the second correlation, by having less shots of their blocked a team would naturally have more SOG. They're better at getting shots through. Additionally if was instead opponents shots that were blocked I'd expect that teams (in general) with higher shots blocked would be higher in the standings. And teams with less, lower.

EDM - 932SB - 5th in the league
PHI - 1067SB - 9th in the league
WPG - 1094SB - 13th in the league
MTL - 1184SB - 7th in the league
HAM - 1212SB - 15th in the league

There isn't a nice easy correlation here. But all I can say looking at that is SB is not a (good) indicator of league position.

I never considered looking at overall team/player SB and correlating it to other stats. So you're stuff is super cool. My only goal was to figure out the descrepency between SOG, and SA (or total shot quality). There's definerly more work to be done here.

*puolivalmiste* · 05-18-2025, 11:33 AM

Could index be showing blocked shots and the box scores shot attempts that were blocked?

To figure that out you could look at last days sims (22/2 - 1/3 in game) for forwards of Minnesota and Toronto and compare if total SBs counted from box scores add up to 29 (MIN) and 28 (TOR) that we got from our hattrick counter which takes those values from index

**bbjygm** · 05-18-2025, 11:33 AM

I'll have to read this further at a later time when I'm less busy, but I doubt my defensive defenseman with top blocks is a volume shooter

***luke*** · 05-18-2025, 12:01 PM

05-18-2025, 11:33 AMpuolivalmiste Wrote: Could index be showing blocked shots and the box scores shot attempts that were blocked?

To figure that out you could look at last days sims (22/2 - 1/3 in game) for forwards of Minnesota and Toronto and compare if total SBs counted from box scores add up to 29 (MIN) and 28 (TOR) that we got from our hattrick counter which takes those values from index

https://docs.google.com/spreadsheets/d/1...sp=sharing

The summation of shots blocked vs the index total are the same

***luke*** · 05-18-2025, 12:11 PM

To Put a bow on this. We have an actual answer by the Dev 3 months ago when someone else pointed this issue out.

https://forums.ootpdevelopments.com/show...ts+blocked

[Image: QnTxMI9.png]

There is still an issue with Shot Quality though. since they are counting shots your team blocked. Instead of shots the other team has blocked

Kami1931 · 05-18-2025, 12:36 PM

What Luke found fits my data and icewindoo's. It is a better answer than I had.

For anyone working with shot quality, be aware that there will be shot quality from opposing teams in your data.

I'm trying to think of a way to strip it out, but it might not be possible.

Navigation

Extra Menu

About us