- Jul 2, 2011
- 1,540
- 1,276
AlrightWhy are you posting charts that don’t really matter? What do you think you’re telling me with them?
I’ll start with the first one. It’s a micro stats chart that samples only 11 games and compares that sample to 185 other dmen’s similarly small samples. The only way you can get to 185 defenseman is including everyone from #1 dmen to 6/7 dmen. What possible value could that info hold with such flawed underlying metrics? I don’t want to bash the work Corey does because I know it’s very time consuming. But it doesn’t change the limited and flawed nature of it.
The 2nd one is an Even strength and Power Play RAPM chart. I’m talking about his usage 5v5. It’s like you wanted to pile on bashing Joker but didn’t pay attention to the details.
The analytics community in general, and Sabres twitter in particular, deserve a smack for convincing people these charts mean more than they could possibly tell us.
Joker’s 5v5 underlying stats per NST.
CF% ——> 51%
SF% ——> 48%
SCF% —-> 50%
HDCF% -> 47%
GF% ——-> 45%
xGF% ——> 49%
OZS% —-> 48%
I generally don‘t use GF% as an underlying metric but as a benchmark to compare the other underlying numbers against (Are they under or over performing their numbers). But I included it here. That looks like a guy who did about as well as could be expected relative to who he is as a player and in the role he was tasked with. I can certainly say he came pretty close to breaking even. But “breaking even“ or “coming close” doesn’t really matter to the point I was making.
A good bottom pairing dman plays about 15mins atoi 5v5. You’d want at least 14+mins.
-Boosh averaged 12:43 for the year and played 9mins or so during the most important games at the end of the season.
-Joker averaged 18:15 a night and played 20+mins during the most important games at the end of the season.
The idea Joker is behind Boosh is disconnected from reality.
I will preface this by saying that I tend to think people overstate the impact of the underlyings for the most part. I've been in many debates about this, and it just comes down to the fact that the public tracking data for the NHL is awful - for pretty much everything (including faceoffs, hits, xgf etc). So, I'm with you there. Also, I want to clarify my position on Joker. I agree that he's probably better off on a third pairing role - and by saying that he's not moving Boosh out of that spot isn't a knock on Joker, but Boosh brings something to our lineup that very few other players do. Jokiharju is best suited for a role that doesn't currently exist on the Sabres defense pairings. Anyways, onto your post.
1. You tried to discredit Jokiharju's RAPM chart because I "didn't pay attention to the details" because it's an EV chart and not 5 on 5. Guess how many minutes Joker played at EV that wasn't 5 on 5? 35 minutes of his 1130:46 total. I don't think that's a big enough impact to suddenly change his charts. Of all the underlyings - I actually like RAPM the best because it's one of the few charts that takes into consideration deployments (I still think there needs to be more done when it comes to on-the-fly starts, considering they're not at all created equal). Anyways, it shows him being below average pretty much across the board.
2. I don't disagree that Corey has a fairly small sample size - however again, I don't anticipate the numbers would drastically change. We can look at pretty much any data point available to us and it's going to show us that Jokiharju isn't that great defensively. I have access to the highest plans of all the major analytics websites/Patreon's and I don't believe there is one thing I could try and twist to show you that he's an even average defenseman. Corey (accurately) points out that he's particularly bad at zone entry denials - that passes the smell test to me.
3. Pointing out his underlying raw numbers isn't really doing much - especially if you're going with "he's basically breaking even" as your argument. He's below average at pretty much everything except CF% - on a team that was (surprisingly) above average in CF% as a whole - so he's actually a negative when it comes to relative rates. Again, every single data point is telling us he's below average.
I do agree that him being 28th in TOI at 5 on 5 is silly. We do need our bottom pair to play more and to roll a more balanced three defensive pairs, but it still leaves Jokiharju in no-mans-land for me. Are we going to bench Boosh for Joki, removing one of a handful of players that actually brings a physical element to the game, and perhaps our only defenseman that can somewhat clear the front of the net? He just doesn't have a place here.