[CP Story] Stat Hack: Is David Rittich the Flames Weakness?

[CorsiHockeyLeague]

My guess is that you have to play deeper in the crease on the PK because the puck is moving coast to coast more, and that doesn't jive with Rittich's "let's challenge everything as much as possible" style. In Smith's case the reason is just that he's old and doesn't have the reflexes he used to.

[IliketoPuck]

Originally Posted by Bingo

It's a bit worse than that.

The 15th best starter has a .914 save percentage. Rittich hasn't been at that level in any of his last three five game segments which totals half of his season.

Some silver linings for me though ...

1) If you told me Smith would be brutal this year I would have gladly taken Rittich at a middling goaltender as a stater (where he's trending)
2) Rittich has some interesting splits ... great at even strength but not when down a man which suggests to me the Flames have a systemic PK issue possibly.

Out of 42 goalies that have faced 100 minutes shorthanded Calgary has the 31st and 38th ranked goaltenders in terms of save percentage.

Rittich has the fourth worst dSV% shorthanded of the 42.

Why?

Haynes had an interesting article on that the other day, that seems to corroborate your observation that it's more of a systemic Flames PK structure type thing than either goalie.

https://theathletic.com/799384/2019/...-historically/

[GranteedEV]

Originally Posted by Bingo

Rittich has the fourth worst dSV% shorthanded of the 42.

Why?

Going by shorthanded CA/60 the Flames have the 4th most aggressive PK in the league holding opponents to 83.13 CA/60. I think it stands to reason that when there are breakdowns they will probably be worse, and I would also be hesitant to look at SH% SV% as something that can be compared with something like dSV% because you're talking about Laine types being left wide open in areas that may not be considered dangerous by typical stats after a PK goes a full minute+ of not allowing a SOG. If the Flames PK were more passive it might boost the SV% but maybe then we don't lead the league in shorties scored.

We've also had a pretty inexperienced PK overall as Andersson, Valimaki, and Hanifin have all been killing for us and none of them have been PKers in the NHL before this year. This goes back to mistakes being magnified playing an aggressive style.

Overall though what we know about special teams goal stats is that the sample size of even one whole season is too random to be useful. 5v5 SV% is the more repeatable/predictable stat.

[Flash Walken]

Want to win with league average goaltending? Easy, score the most goals of any team in the league:

Originally Posted by Flash Walken

definitely the exception, but some food for thought:

The year Chicago won their first cup in 09/10, 34 year old Cristobal Huet put up an .895 sv% over 48 games while Niemi sported a .912 sv% playing 39 games. That is roughly the split Smith and Rittich have played thus far.

Even with his poor numbers, Huet managed 4 shutouts. Mike Smith is also essentially on this pace despite his .888 save percentage.

The first half of the season, Huet was having a much better season than Smith has thus far, but he began to falter in the second half of the year while Niemi stayed his consistent self.

Huet eventually lost the starters net to Niemi that spring, and was able to raise the cup after sporting a .910 and 2.63 stat line in 22 games in the playoffs.

How do you win a cup allowing 2.63 goals against per game with a .910 save percentage?

You score 3.55 goals per game, the highest goals per game of any cup winning team between 2010 and last year when the capitals won with 3.58 goals per game.

Chicago finished 3rd that year in the regular season with 3.20 goals per game.

The Flames right now are sitting at 3.53, 6th in the league.

Chicago finished 4th best in goals against that season with 2.48. Right now, the flames are 7th best with 2.78

It's an outlier, but it's definitely possible if you can score enough. If the flames had another scorer, they could be that team.

This whole goalie conversation can be registered as largely inconsequential with the addition of two more forwards that can score.

[Bingo]

Originally Posted by GranteedEV

Going by shorthanded CA/60 the Flames have the 4th most aggressive PK in the league holding opponents to 83.13 CA/60. I think it stands to reason that when there are breakdowns they will probably be worse, and I would also be hesitant to look at SH% SV% as something that can be compared with something like dSV% because you're talking about Laine types being left wide open in areas that may not be considered dangerous by typical stats after a PK goes a full minute+ of not allowing a SOG. If the Flames PK were more passive it might boost the SV% but maybe then we don't lead the league in shorties scored.

We've also had a pretty inexperienced PK overall as Andersson, Valimaki, and Hanifin have all been killing for us and none of them have been PKers in the NHL before this year. This goes back to mistakes being magnified playing an aggressive style.

That still doesn't explain the differential in my opinion.

Five on five:
Flames allow the 5th least CA/60
Flames allow the 9th least SCA/60
Flames allow the 19th least HDCA/60

Shorthanded:
Flames allow the 4th least CA/60
Flames allow the 12th least SCA/60
Flames allow the 16th least HDCA/60

The differentials are pretty consistent five on five (where he excels) to shorthanded (where he gets lit up)

[GranteedEV]

Originally Posted by Bingo

That still doesn't explain the differential in my opinion.

Five on five:
Flames allow the 5th least CA/60
Flames allow the 9th least SCA/60
Flames allow the 19th least HDCA/60

Shorthanded:
Flames allow the 4th least CA/60
Flames allow the 12th least SCA/60
Flames allow the 16th least HDCA/60

The differentials are pretty consistent five on five (where he excels) to shorthanded (where he gets lit up)

You're also comparing a sample size of 668 shots in 1348 minutes (the equivalent of 22.4 NHL games)

to

a sample size of 108 shots in 147 minutes (the equivalent of two and a half NHL games)

[FlamesAddiction]

Originally Posted by Flash Walken

Want to win with league average goaltending? Easy, score the most goals of any team in the league:



This whole goalie conversation can be registered as largely inconsequential with the addition of two more forwards that can score.

I would be curious to know the data on when Chicago scored and if they scored first more of the time.

The Flames seem to have poor starts and then turn it on in the 3rd. It's worked so far in the regular season, but scoring typically becomes harder if you are playing from behind because the other team can focus more on defense. If our goalies keep letting in 1st period muffins, it doesn't help.

[Bingo]

Originally Posted by GranteedEV

You're also comparing a sample size of 668 shots in 1348 minutes (the equivalent of 22.4 NHL games)

to

a sample size of 108 shots in 147 minutes (the equivalent of two and a half NHL games)

You brought up the CA60 at PK, not me.

I'm just saying there isn't really a difference between the rates in rank five of five vs shorthanded.

Doesn't explain Rittich's difference at all.

[The Fonz]

Rittich has the 2nd best win% in the league (>= 25GP), and the Rittich:Smith splits for SV% and GAA are .915:.889 and 2.55:3.11

Is David Rittich the Flames weakness? No.

[DeluxeMoustache]

Bingo, your article starts by using xSv% and dSv%, etc. They attempt to improve on sv%, by attempting to account for distribution of shot quality (now the buckets they place shots in are a fine start, but still not where they need to be)

Moving to then compare straight ahead sv% in game chunks is understood, and it makes sense to want to compare chunks of games, but it is still bit of a step backwards. I couldn’t see on the basic Corsica interface how to isolate chunks of games, unfortunately.

PK save percentages are going to be affected largely by team defensive play. The team with man advantage often has the luxury of waiting for their best shooting opportunity and no two teams’ PKs are equal.

Is there some correlation between the Flames success in scoring shorties, and a lower sv% on the PK, in that maybe the PKers are willing to cheat a bit to the offensive side and expose themselves a bit? Would the stats you are relying on capture any positional cheating and resulting exposure?

At some point, it comes down to true shot quality, and the proxies and data are not yet where they need to be.

People have to decide if they like Rittich or not.

The one good thing, for me, is that Rittich clearly does not obsess over the very rare bad goal and let it affect him like it seems to affect some people on CP

[icecube]

I'm a member of the Rittich fanclub. I think he has what it takes to be a starter. But I would love a guy like Talbot to come share the load and have a play the hot hand situation. I think the Flames would too and I bet they are trying to get their hands on him or someone similar as we speak.

[Bingo]

Originally Posted by DeluxeMoustache

Bingo, your article starts by using xSv% and dSv%, etc. They attempt to improve on sv%, by attempting to account for distribution of shot quality (now the buckets they place shots in are a fine start, but still not where they need to be)

Moving to then compare straight ahead sv% in game chunks is understood, and it makes sense to want to compare chunks of games, but it is still bit of a step backwards. I couldn’t see on the basic Corsica interface how to isolate chunks of games, unfortunately.

PK save percentages are going to be affected largely by team defensive play. The team with man advantage often has the luxury of waiting for their best shooting opportunity and no two teams’ PKs are equal.

Is there some correlation between the Flames success in scoring shorties, and a lower sv% on the PK, in that maybe the PKers are willing to cheat a bit to the offensive side and expose themselves a bit? Would the stats you are relying on capture any positional cheating and resulting exposure?

At some point, it comes down to true shot quality, and the proxies and data are not yet where they need to be.

People have to decide if they like Rittich or not.

The one good thing, for me, is that Rittich clearly does not obsess over the very rare bad goal and let it affect him like it seems to affect some people on CP

Here is the Tampa PK heat map for the PK. They're ranked number one in PK success, but also number one in expected save percentage (tabulation of what they give up).

https://hockeyviz.com/fixedImg/shotL.../1819/T.B/wrap

Calgary is ranked 19th in PK success, and 15th in expected save percentage shorthanded.

https://hockeyviz.com/fixedImg/shotL.../1819/CGY/wrap

Is it me or does Calgary's map look less dangerous?

[DeluxeMoustache]

Originally Posted by Bingo

Here is the Tampa PK heat map for the PK. They're ranked number one in PK success, but also number one in expected save percentage (tabulation of what they give up).

https://hockeyviz.com/fixedImg/shotL.../1819/T.B/wrap

Calgary is ranked 19th in PK success, and 15th in expected save percentage shorthanded.

https://hockeyviz.com/fixedImg/shotL.../1819/CGY/wrap

Is it me or does Calgary's map look less dangerous?

Interesting. Haven’t seen these before

The dark green around the crease suggests TB looks to be very good at boxing guys out. You don’t appear to have a lot of allowance for guys like Matthew Tkachuk!

[CSharp]

Rittich is still able to shut the door in the 3rd periods. The team needs to start out better and play better in their own zones. The last few games after the All-star break, the team has given up weak goals - all within the first periods. Not all are any of the goalies' fault.

[Sylvanfan]

How much do some of these last minute goals given up with two goal leads drag down Rittich's save percentage?

Not so much in the past 5 game segment as in the previous one. I would think that if you were to take out 3 goals from those 3 straight games the numbers there look really good.

Goals against are goals against, but there is a timeliness element to when goals are given up too. You don't like allowing any goals, but if you have to give up a goal...better to do it with 20 seconds left and your team is up two goals type of thing.

The number of goals where the opponents have an empty net, might also shed light into the PK as well, and how the Flames defend when down a skater.

I still think it's a short sample size with Rittich, but I'm be willing to give him the net but as others mention, make it so that he's starting 5 out of every 7 games in these last 20 type of thing to manage the work load and see what you have. I still think they have to look for backup options, even to the point of ensuring the #3 depth option is better than what they currently have.

Realistically if you can get that Chris Osgood type goaltending, the Flames can likely go a long ways with it.

[ResAlien]

It’s not rocket appliances. Rittich has looked great all year except for giving up crappy short side high goals for a bit (which was corrected) and a stretch of giving up relatively meaningless last minute goals in multi goal lead games. He has definitely looked shaky since the break. As long as he can get back to what he was before the break he looks to be a bonifide starter, middle of the pack keeper at worst. The hope is he does that soon.

I’d love a similar stat dive to the OP on our backup. I’d wager that may be enlightening.

[Bingo]

Originally Posted by ResAlien

It’s not rocket appliances. Rittich has looked great all year except for giving up crappy short side high goals for a bit (which was corrected) and a stretch of giving up relatively meaningless last minute goals in multi goal lead games. He has definitely looked shaky since the break. As long as he can get back to what he was before the break he looks to be a bonifide starter, middle of the pack keeper at worst. The hope is he does that soon.

I’d love a similar stat dive to the OP on our backup. I’d wager that may be enlightening.

https://www.calgarypuck.com/2019/01/...th-the-backup/

[ResAlien]

Originally Posted by Bingo

https://www.calgarypuck.com/2019/01/...th-the-backup/

Thanks for that. Worst or second worst backup in the league, basically what I expected. Definitely puts more pressure on Rittich to keep this team afloat knowing the backup is so poor. Maybe that plays to the mental side of goalie, as the pressure on the season mounts knowing if you don’t do it the backup can’t?

[Textcritic]

Originally Posted by ResAlien

Thanks for that. Worst or second worst backup in the league, basically what I expected. Definitely puts more pressure on Rittich to keep this team afloat knowing the backup is so poor. Maybe that plays to the mental side of goalie, as the pressure on the season mounts knowing if you don’t do it the backup can’t?

I suspect that pressure is mitigated some by recognizing that for all his flaws Mike Smith still wins more games than he loses. Rittich and Smith are purportedly close—that likely counts for a lot more for him than the numbers Smith posts.

[Bingo]

And numbers in a similar format to Rittich piece ...

Segment	SV%	QS%	RBS%	Wins
1	0.889	40%	60%	3
2	0.853	20%	40%	2
3	0.915	60%	0%	2
4	0.911	60%	20%	3
5	0.891	40%	20%	3

Terrible start ... first two segments were brutal with awful save percentage, and too many Really Bad Starts

Then two segments with an uptick in save percentage, and far fewer ugly games.

A step back in the last five games to what we saw to start the year.

Should be pointed out that Smith's segment 3&4 are better than Rittich segment 4&5 .. that statement should get me yelled at I'm guessing