I've been thinking about how to put together something that will be both informative and accurate. Statistically coaching is difficult to nail down.
Let's first look at those statistics that are very much based upon team strategy as strategy falls soley under the purvue of coaching:
Powerplay:
A general look at the Sharks powerplay during the tenure's of the last 3 head coaches. I am leaving the 02/03 season out of the average by the way, because we had two coaches that season:
Clearly Todd's powerplay strategy is working. He was known as the architect of the powerplay in Detroit as well, and clearly excels. The quality of the team plays a huge part in this as well though, so those numbers need to be looked at as at best... noteable.
Penalty Kill
Again a look at the PK under 3 different head coaches
Year
Coach
%
Rank
97-98
Sutter
85.1
12
98-99
Sutter
85
13
99-00
Sutter
83.6
16
00-01
Sutter
86.1
5
01-02
Sutter
85.8
9
02-03
Wilson/Sutter
81
30
03-04
Wilson
85.2
7
05-06
Wilson
80.7
23
06-07
Wilson
83.3
14
07-08
Wilson
85.8
1
08-09
McLellan
83.3
5
09-10
McLellan
85
5
10-11
McLellan
79.6
24
11-12
McLellan
77.5
28
DS AVG%
Rank
RW AVG%
Rank
TM AVG %
Rank
85.1
11
83.75
11
81.35
16
As you can see the PK rank under McLellan's regime has been clearly the worst even though the Sharks arguably had the best roster during these years. Oddly, it has seemingly gotten worse the longer Todd has been the head coach.
Shots Again Per Game:
A look at how many shots allowed per game, per year:
Year
Coach
SA/G
Rank
97-98
Sutter
24.6
4
98-99
Sutter
27
10
99-00
Sutter
28.7
19
00-01
Sutter
26.5
9
01-02
Sutter
28.5
21
02-03
Wilson/Sutter
29.2
21
03-04
Wilson
28.9
20
05-06
Wilson
26.6
4
06-07
Wilson
26.2
3
07-08
Wilson
24.2
2
08-09
McLellan
27.2
1
09-10
McLellan
31.4
20
10-11
McLellan
28.9
8
11-12
McLellan
28.3
6
DS avg sa/g
Rank
RW avg sa/g
Rank
TM avg sa/g
Rank
27.06
13
26.5
7
28.95
9
Sharks are allowing more shots per game than under any other head coach.
Shots Per Game and Goals Per Game:
A Look at the Sharks shots for per game and goals per game:
Year
Coach
S/G
Rank
G/G
Rank
Conv%
97-98
Sutter
26.5
22
2.56
18
9.7
98-99
Sutter
26.1
22
2.39
21
9.2
99-00
Sutter
27.4
13
2.74
16
10
00-01
Sutter
27.6
15
2.65
17
9.6
01-02
Sutter
27.2
18
3.02
4
11.1
02-03
Wilson/Sutter
28.1
16
2.61
15
9.3
03-04
Wilson
28
17
2.67
11
9.5
05-06
Wilson
30.3
11
3.23
7
10.6
06-07
Wilson
28.6
21
3.12
6
10.9
07-08
Wilson
29.6
11
2.63
19
8.8
08-09
McLellan
33.2
3
3.06
7
9.2
09-10
McLellan
31.7
6
3.13
4
9.9
10-11
McLellan
34.5
1
2.96
6
8.6
11-12
McLellan
34.2
1
2.70
12
7.9
DS avg S/g
Rank
RW avg s/g
Rank
TM avg s/g
Rank
27
18
29.13
15
33.4
3
DS avg g/g
Rank
RW avg g/g
Rank
TM avg g/g
Rank
2.67
15
2.91
11
2.96
7
DS avg conv
RW avg conv
TM avg conv
9.9
10
8.9
So this is very interesting. Todd is clearly having a significant effect on the shots on net, but a statistically insignificant effect on goals per game. His teams conversion rate is much lower than either previous coach, and trending down sharply.
Under Todd McLellan the Sharks are having a MUCH more difficult time putting the puck in the net. They are having to do a lot more work to equal the same number of goals per game. They are working harder, not smarter.
I hope this is somewhat enlightening to people. By the numbers Todd is not a bad coach by any means, but during his tenure the Sharks have gotten significantly statistically worse defensively and on the PK. They are putting about the same number of goals per game, but having to put up a lot more shots in order to do so. At best the Sharks are equal offensively under Todd (with better personal arguably) and debatable less effective. The only category where they have clearly improved is the PP, and that more than any other stat is impacted by the players themselves.
Last edited by hockeyball: 03-05-2012 at 05:38 PM.
Thanks for the breakdown, yeh TMac does some things well, I wonder how much of that 08-09 start was him coming in and how much was just a well balanced team firing on all cylanders. He really hasn't done anything amazing since then.
I believe it has alot to do with the new league stance on active interference, allowing teams like Blues and Detroit to create a very big defensive counter to Sharks' offensive playstyle. Sharks take way too long to get out of their own end and the opponents can just line up the interference wall on their blue line. Getting Havlat back should help with the speed a bit, but something big needs to be done to the offensive zone entry.
Thank you for the effort HB. Very enlightening and supports what many of us have said. It is a strong counter to the pure possession argument.
Coaches acknowledge the biggest issues for coach accountability are PP and PK. This is the area with the widest variation in coaching strategy.
On PP, it is in part personnel. The PPs that standout are the ones with both a strong blueline QB (eg Boyle, Datsyuk) and a strong forward QB (eg Thornton, Lidstrom). A team can have the personnel but relatively fail as well. There is a difference between a 19% PP and a 23% PP (coaching difference). Without the personnel, you are frequently talking about 16% and under. It is interesting that Nashville is going with Suter/Weber as QBs without the real forward component. Very unique and kudos to Trotz.
My take is that when TM arrived, the Sharks were still leaving the basic PK to the players. They moved to the strategic change in his third year.
I'm know this has been talked about endlessly in the past, but WTF happened between 09-10 and 10-11 in regards to our PK? Manny Malhotra? Perhaps Nabokov was bailing us out more than we thought?
Team shooting percentages have dropped across the board for the past 3 seasons (from 9.4% to 9.1% to 8.9%) so you have to take that into account - this isn't a McLellan thing, it's a league wide, goals are down issue.
Aside from the fact that you're making the two most basic errors of statistical analysis (concluding causation rather than correlation and assuming descriptive historical data has predictive value without doing any statistical testing of the data)...
It's interesting, yet basically meaningless data upon which to draw any conclusions.
Team shooting percentages have dropped across the board for the past 3 seasons (from 9.4% to 9.1% to 8.9%) so you have to take that into account - this isn't a McLellan thing, it's a league wide, goals are down issue.
Also, is there a reason you left out GA/G?
Goalie has too big an impact on GA/G for it to be usefull as a coaching statistic. I considered it, but felt it wasn't particularly helpful.
I believe it has alot to do with the new league stance on active interference, allowing teams like Blues and Detroit to create a very big defensive counter to Sharks' offensive playstyle. Sharks take way too long to get out of their own end and the opponents can just line up the interference wall on their blue line. Getting Havlat back should help with the speed a bit, but something big needs to be done to the offensive zone entry.
Holy crap that was painful to watch against the Blues. I swear we played 75% of the game in our own zone.
Also, I was surprised when they finally called interference on Colaiacovo. I have resigned to the fact that we dont get any interference calls in our favor anymore.
__________________
What?! Look, he thinks he's people!
But you think a teams shooting percentage is a useful coaching stat?
I'm honestly not sure what you are getting at.
A team shooting percentage takes a lot of players (large sample) against a lot of goalies (large sample) over an entire seasons (large sample) to get a number.
A GA stat takes a lot of players (large sample) against two goalies (small sample) over an entire season (large sample).
So you can see one of those two has a significant bottleneck. If your goalie sucks (or is great) it's going to have a large impact on your GA. With a shooting percentage stat, no one player can have a significant impact.
Holy crap that was painful to watch against the Blues. I swear we played 75% of the game in our own zone.
Also, I was surprised when they finally called interference on Colaiacovo. I have resigned to the fact that we dont get any interference calls in our favor anymore.
Interference isn't called for any team after the second half of the season. It's a coaching error that we aren't taking full advantage of it, look at the teams who are in the tops of the standings and they are teams that are capable of running good picks. Feels like the coaching staff has an attitude where it's poor gamesmanship but if refs are going to let it go we are going to have to get better at it.
A team shooting percentage takes a lot of players (large sample) against a lot of goalies (large sample) over an entire seasons (large sample) to get a number.
A GA stat takes a lot of players (large sample) against two goalies (small sample) over an entire season (large sample).
So you can see one of those two has a significant bottleneck. If your goalie sucks (or is great) it's going to have a large impact on your GA. With a shooting percentage stat, no one player can have a significant impact.
Team shooting percentages have dropped across the board for the past 3 seasons (from 9.4% to 9.1% to 8.9%) so you have to take that into account - this isn't a McLellan thing, it's a league wide, goals are down issue.
Interesting on SH%. Number of PPs has fallen off and I am sure that PP SH% is better generically than 5on5 SH%. That might be the causation on league #'s.
There is also an issue of leaguewide style of play. It is clear that possession is getting more emphasis from coaching commentaries.
It shouldn't be any different than using G/G as a coaching statistic.
Goals per Game is a direct result of offensive strategy and again is a lot of players, against a lot of goalies, over a full season. No one player can have a significant effect on the statistic.
If for instance the Sharks had 10 goalies they rotated around, goals against would be a lot more reliable to use as a strategy analysis.
Seriously, if you guys want to list it, go ahead. I looked at it, it doesn't really add anything to the conversation in my opinion, but you are welcome to add it here.
With a shooting percentage stat, no one player can have a significant impact.
No, but a lot of factors that have nothing to do with coaching can have an impact and you haven't in any way eliminated/calculated their effects, so the descriptive historical statistic is totally meaningless in terms of drawing any conclusions about anything.
I disagree, a coach designs the offensive strategy. If a team is not using net front presence, effective passing plays (to get the goalie out of the way) etc that is going to have a huge effect on conversation rate.
As an example. If team A takes 10 shots by skating into the zone and putting the puck into the goalie's pads and Team B skates into the zone and executes crisp passing plays designed to throw the goalie out of position, regardless of the personnel team B is likely going to have a higher conversion rate.
Goals per Game is a direct result of offensive strategy and again is a lot of players, against a lot of goalies, over a full season. No one player can have a significant effect on the statistic.
It's also the result of the talent on the team, year by year; talent on other teams, year by year; changes in rules/refereeing year by year. You are drawing a conclusion without even developing a model and testing its signficance. Again, it's interesting historical data, but it has no predictive value whatsoever unless you build a model and test it.
