Saturday, January 31, 2009

The Effect of Back to Back Games

A couple of days ago I wrote about one reason that's been given by fans and media for a team playing below their standard level of ability. In that study I concluded that there is no negative effect on a team in their first game back after a long road trip, and that, if anything, there is a slightly positive effect. That particular myth is not backed up by a sober look at the real wins and losses by actual NHL teams. But what about the effect of playing in back-to-back games? I remember playing two or three games a day in tournaments when I was twelve years old, and while I may have felt tired by the end of that last game, I was twelve and we're talking about professional athletes here. So does playing on back to back nights actually effect the ability of a team or is it yet another aspect of Canadian hockey mythology with no roots in reality?

In order to do this study I only used instances where the second half of the back to back was on the road so that I could use the road record overall as a control. After last night's games there were a total of 188 games where the road team was playing the second half of a back to back. Given that there are only 729 games in total that represents just over 25% of all games. I was shocked that fully a quarter of games are played with the road team on the back half of a back to back.

Anyway, the good news about that is the sample size for this study is probably large enough to come to a conclusion. In the aforementioned 188 games the goal differential (not including shootout goals) for the road team is -124 or -0.660 goals per game. The points percentage of the road team is 44.9% and the winning percentage is 38.8%. In the other 541 road games that teams have played this year the points percentage is 51.1% and the winning percentage is 46.0%.

The difference is huge. In back to back situations teams are much less likely to register points (a difference of 6.2%) and the winning percentage falls dramatically (a difference of 7.2%, the same as the gap between the St. Louis Blues and the Anaheim Ducks). This meme is definitely based in reality which of course leads to the question of why on earth 25% of NHL games are played with the road team at a significant disadvantage. The best answer I can think of is that it makes it more likely for the fans to be sent home happy. Another thing this may point to is a possible effect on playoff games. If there are fewer back to back games in the playoffs the road team has a much better chance to win than they do in the regular season.

Friday, January 30, 2009

Oilers v. Wild - Titus 1:6-9

Our captain must be blameless, not chasing after puck bunnies, a man whose children believe in the Oilers and are not open to the charge of being a fan of the Wild or Flames. Since this captain is entrusted with leading God's favoured team, he must not be open to any charge—not overbearing on his young teammates, not quick-tempered on the ice, not given to drunkenness in front of a camera, appropriately violent, not chasing after wealth. Rather he must be hospitable to teammates, one who loves winning, who is self-controlled, ready to take criticism, spending time in the community and even more working on his game. He must hold firmly to the trustworthy message that God loves Edmonton so that he can encourage others with this sound teaching and beat back those who oppose it.

Thursday, January 29, 2009

First Game Back After a Long Road Trip II

In November I wrote a post about the effect of coming home after a long road trip. At the time the Oilers had just lost with some uninspired play against the Maple Leafs. Given that the Oilers were just blown out 10-2 by Buffalo, I was reminded of that previous post and thought that I might update the findings since the sample size will now be much larger. My hypothesis is that fans like to think of reasons why a team might underperform or otherwise "come out flat." One of the reasons generally accepted in Canadian hockey mythology is that of "the first game back after a long road trip."

In November, the results showed that the home team had a points percentage of 63.5% in the first game back after a long trip (the trip needs to be three consecutive road games or more) and a points percentage of 59.4% in all other situations. The results were very preliminary because at the time there was a total of 320 games and only 37 of those games qualified under the "first game back" criteria.

There has now been a total of 715 games, 93 of which meet the criteria for "first game back." In the 93 games that qualify the home team has a points percentage of 64.0%, a slight increase from November. The goal differential in these games (excluding shootout goals) is +42 (297-255) or +0.452 per game. In the other 622 games the home team has a points percentage of 61.4%, a much larger increase, but still 2.6% behind. I think that this sample is large enough to conclude that there is certainly no negative effect from coming home after a long trip and that there may be a slightly positive effect when the team first gets home.

Now the easiest objection is that three games is not enough to constitute a long road trip, which may well be fair. As such, I have collect those situations where a team was coming back from a trip of 5 games or more. I think that it is pretty certain that this should constitute a long road trip, but there are still only 23 games that qualify here, so not all of the teams are even represented. Nonetheless, the results in those 23 games have the home team achieving a points percentage of 63.0% with a goal differential of +20 or +0.870 per game. The points percentage has slightly decreased, but the goal differential has close to doubled. These disparities are probably the result of a small sample size. In the other 692 games the points percentage is 61.7%, so still over 1% behind. These last numbers are still too small to come to a firm conclusion but they do support the conclusion reached above: there is no negative effect on the home team in their first game back after a long trip. If anything, there is a slightly positive effect.

Perhaps I'll check back at the end of the season to see if these conclusions, especially with regard to road trips five games or longer, still hold up.

Wednesday, January 28, 2009

Scoring Chances - Games 1-41

Now that we've broken down the scoring chances into smaller ten game segments it would, of course, also be instructive to see what's happening in the macro view (especially important when the alternative micro view is a 10-2 loss). Once again, thanks very much to Dennis for tracking the scoring chances and if anyone hasn't yet seen his work or that of Bruce who is expressing this data in per 60 rate, then you can find all of that here.

(For those unfamiliar with the metric a player is awarded a chance any time someone on the ice has a chance to score. He is awarded with a “chance for” if someone on his team has a chance and awarded with a “chance against” if someone on the opposing team has a chance to score. The results are broken down into three game states, EV (even strength), PP (Power Play) and SH (Short-Handed). The players are organized according to their jersey numbers.)

Games One to Forty-One, October 12 to January 11
Missing Games 22 at St. Louis, 33 at Vancouver and 40 vs San Jose


Now there are other things to consider like Quality of Competition that need to have some bearing on the analysis. It's safe to say that for the most part Horcoff, Hemsky and Penner have been thrown into a Power v. Power relationship in which they've been taking on most of the hardest minutes on the team. All three have a positive differential at EV and all three have been contributing positively on the first unit PP. That line has, quite simply, been delivering results even if their offence may be a little shy of where fans had hoped in the beginning of the year.

Cole received some tougher sledding with Horcoff and Hemsky in the first ten games and Reddox received his share with Horcoff and Penner in the last five or six games while Hemsky was injured. Neither of those two had particularly good numbers in that role but Cole and Gagner were especially dominant in the last ten games bringing him into positive territory. It also vaults Gagner (hopefully he's not out for too long) into the lead in scoring chance differential among forwards with a +24. Gagner may not be putting up as many points as I had expected, but, much like the top line, he has been delivering very positive results. The kid is only 19 years old too. It seems very likely that he will be a difference maker for years to come.

Reddox, meanwhile is a fourth liner who has been able to break even in that role but clearly isn't ready for primetime on the top line. He's also not outscoring on the fourth line and, at this point, probably shouldn't be considered better than a tweener. Stortini is in the negative here, as is Brodziak, though in Brodziak's case the split between offensive and defensive zone faceoffs is extreme and, as such, understandable.

So if the fourth line can, mostly, is only down slightly, the first line is rocking and the second line can be dominant, the problem areas are now somewhat obvious. Ethan Moreau and the third defensive pairing (Staios with Smid or Strudwick) have been completely terrible. Just awful. Given that in the case of Staios and Moreau they're overpaid for bottom roster players, one would hope that this would be an area of strength. That is, quite simply, not the case.

Speaking of Staios and Moreau one of their primary responsibilities on this team is killing penalties. The penalty kill has been terrible. That much is sure. Given the results those two have at even strength I am currently in favour of trying others in their spots and removing them from these assignments entirely. If it helps, that's an easy fix to the problem. If not, well then the problems run much deeper. The Oilers have allowed about (adding together all the individual chances against and dividing by four) 173 chances against while creating about (adding together all the individual chances for and dividing by five) 152 chances for while on the power play. We also know that over the course of the season the Oilers have scored (not including the three games mentioned above) 30 power play goals for a rate of one goal every 5.07 chances. On the PK they've allowed 42 goals for a rate of one goal every 4.12 chances. So the Oilers are not only allowing more chances, but they're also stopping chances at a much lower rate. Whether this is simply the quality of the goaltending, or also the quality of the chances, I'm not sure, but this gap needs to close if the Oilers hope to win the special teams battle in the second half.

Tuesday, January 27, 2009

Oilers v. Sabres - Luke 16:1-14

Then Kevin told this story: "There was once a rich group of men whose manager was accused of wasting their possessions. So they called that manager in and asked him, 'What is this we hear about you? Give an account of your management or you may not be our general manager any longer.'

Then I said... um... the manager said to himself, 'What should I do now? My master is threatening to take away my job. I'm not young enough to play, and I'm ashamed to beg for an arena...' After much thought I said... er... he said to himself, 'I know what I'll do so that, when I lose my job here, people will welcome me into their houses!'

So he called in two slaves from two different masters. He asked the first, 'How much does your master owe you?' 'They have offered only three and a half million over three years, playoffs included!,' he replied. So I told him, 'Sit down quickly, and make it seven million over seven years!'

Then he asked the second, 'And how much are you owed?' 'The offer is about two million,' he replied. So I told him, 'Take your contract and make it four million!'

When Cal and the boys heard this they commended the dishonest manager because he had acted shrewdly. Inspired by his shrewdness, they decided not to fire the manager, but to sell their possession! For the people of this world are more shrewd in dealing with their own kind than those who are honest.

So then, use worldly wealth to gain friends for yourselves so that when it is gone, your name might be written on the Cup of a lord. Someone who played well for very little will play even better for much, and whoever is lazy with very little will be even lazier with much. If you have not been trustworthy in handling worldly wealth, who will trust you with the true riches of playoff triumph? And if you have not been trustworthy with someone else's property, who will trust you with a lord's Cup? No servant can serve two masters. Either he will hate the one and love the other, or he will be devoted to the one and despise the other. You cannot serve both the interests of winning Cups and hoarding money."

Thomas Vanek and Dustin Penner, who loved money, heard all of this and were chuckling to themselves.

Monday, January 26, 2009

Scoring Chances - Games 31-41

This is the fourth segment of Oilers scoring chances as counted by Dennis. Bruce has recently put up a very nice series of posts expressing some of the totals in terms of a rate per 60. It's interesting stuff that helps to bring things into focus. I definitely recommend giving it a read. The first three game segments are here, here and here.

(For those unfamiliar with the metric a player is awarded a chance any time someone on the ice has a chance to score. He is awarded with a “chance for” if someone on his team has a chance and awarded with a “chance against” if someone on the opposing team has a chance to score. The results are broken down into three game states, EV (even strength), PP (Power Play) and SH (Short-Handed). The players are organized according to their jersey numbers.)

Games Thirty-One Through Forty-One, December 19 to January 11
Missing Games 33 at Vancouver and 40 vs San Jose

Nilsson manages to get benched for game 32 near the beginning of this segment, which is somewhat surprising given the terrible set of games he had without punishment from 11-20. In fact, it seemed as though he had turned it around from 21-30, posting mostly positive numbers and missing a couple of games due to injury. This leads me to believe that some of the decisions with regard to benching have more to do with issues of personality or attitude than they do actual long-term results. The other option is that MacT is making decisions based on singular events that stand out to him. Regardless, he seems to have responded well upon returning, posting a positive differential until he gets reinjured in game 41.

Hemsky was also injured during this time period and the effects of his injury can be seen in the differential posted by his linemates, Horcoff and Penner. At EV, Horcoff went 30/35 and Penner went 24/28 in their six games without Hemsky during this time. In fairness to them, the sample is quite small, and the three games prior (with Hemsky) they had been struggling already. The offensive slack was picked up by the Cole/Gagner tandem who posted fantastic numbers this segment. I believe Cole's +26 represents the biggest positive differential for an Oiler in any segment this year. The two of them were simply dominant, and if that can continue when Hemsky is back full-time, this team will be very difficult to handle at EV.

The flip side to that is the horror show that is Ethan Moreau. It seems that no matter who he's paired with, the results simply stay brutal. He was paired with Cogliano during this time period and Cogs promptly posted his worst results of the season by quite a margin at -11. The pair were able to create a fair number of chances but were completely unable to limit chances against. Moreau was very smart in signing his long-term deal and very lucky to be captain, because on merit, the man should probably be in the press-box.

The call-ups mostly struggled this segment. Both Brule and Reddox posted brutal numbers, although Reddox spent some time filling in for Hemsky against the toughs. Note: he's not ready. He managed to play to even in the NHL in a lesser role so I could see him staying with the big team if he's able to add some offence against weak opposition, something he had been able to do against the toughs in the AHL.

The PK improved greatly in this segment, thanks mostly to increased roles for Cole and Grebeshkov, as well as improved play from some of the stalwarts on that unit, especially Horcoff. This is really the first segment in which the PK was able to generate significant chances in the right direction, suggesting that the flow of play was more favourable for them at this time. Judging from the PP+ totals posted by the D as compared with the previous segments, the Oilers, surprisingly, were able to create a lot of chances on the PP without Hemsky in the lineup.

Saturday, January 24, 2009

Scoring Chances - Games 21-30

This is the third ten game segment of Oilers scoring chances, as tracked by Dennis here. I think it's interesting to break things up into smaller portions because it enables us to see how certain coaching decisions immediately impact the team, as well as how a prospect might be able to establish himself on the team with strong play early. We saw in the previous segment that MacTavish tends to have quite a bit of time for some things that have worked in the past (the 24-43 pairing) in spite of the "blender" reputation. The first two segments of this exercise are here and here.

(For those unfamiliar with the metric a player is awarded a chance any time someone on the ice has a chance to score. He is awarded with a “chance for” if someone on his team has a chance and awarded with a “chance against” if someone on the opposing team has a chance to score. The results are broken down into three game states, EV (even strength), PP (Power Play) and SH (Short-Handed). The players are all organized according to their jersey numbers.)

Games Twenty-One Through Thirty, November 26 to December 17
Missing Game 22 at St.Louis

Nilsson misses games 24 to 27 and Gagner misses games 25 and 26, both due to injury. Both Nilsson (in three games) and Gagner (in four) go 15/11 at EV after their injuries. They may have been a blessing in disguise since the ten to fifteen previous games had been a big struggle for both of them. Their absence also gives a chance for some other forwards to get into the lineup. Brule and Sestito both only get a cup of coffee and are sent back down. Schremp's numbers get killed by his 0/5 against the Sharks and he's sent packing after one more game. Reddox continues his break-even play and at this stage he's breaking even after twelve games, establishing a small track record of success.

Staios and Smid provide a good third pairing once MacT finally gets away from the 24-43 combination in game 25. Smid's totals as a forward are 2 GP and 2/3 at EV. Strudwick's moves to forward for 4GP and 2/5 at EV. 24 also leads the team in a nice return to respectability in this segment on the PK, though it's possible that the biggest difference here is simply fewer penalties to kill.

Both on the PP and at EV the 27-10-83 line does very well. Given the results it's hard to figure why MacT ever takes 27 off the first PP unit. I'd guess that the only reason is to reward or punish his effort level because it sure isn't a problem with the long-term results when he's out there.

Scoring Chances - Games 11-20

This is a closer look at the Oilers' scoring chances for games eleven through twenty logged by Dennis here. There is a lot of interesting discussion there under almost every game with a lot of helpful contributions from Vic and Bruce in particular. I would encourage anyone interested in this discussion to head over there throughout the season and join in.

(For those unfamiliar with the metric a player is awarded a chance any time someone on the ice has a chance to score. He is awarded with a “chance for” if someone on his team has a chance and awarded with a “chance against” if someone on the opposing team has a chance to score. The results are broken down into three game states, EV (even strength), PP (Power Play) and SH (Short-Handed). The players are all organized according to their jersey numbers.)

Games Eleven Through Twenty, November 2 to November 20

This ten game segment includes the experimental 18-89-83 line which didn't work out well at all as far as scoring chances are concerned. This may also reflect Horcoff’s importance to Hemsky in his role as an outscorer, although that sentiment may need to be somewhat tempered since he was being replaced by a 19-year-old centerman and the line was also saddled with 18 on the left flank who hasn't had good results with anyone this season.

Penner was pressboxed for games 17 and 18. His EV scoring chances in the two games following were 03/09 which means in the six games prior to his benching he was 24/17. I suppose he must have looked lazy on a few plays but that’s a pretty good six game stretch to be punished for and while the conventional wisdom is that he responded to MacT’s tongue-lashing well, it looks like that had a lot more to do with an immediate increase in time on the PP than an immediate improvement in his overall play. His PP stats are 06/00 in the six games before the benching and 10/01 in the two games after which is entirely a function of his TOI.


Nilsson really struggled during this ten game stretch but his benching doesn’t come until later.

Game 20 is Smid’s first as a forward. The numbers here make it jawdropping that MacT didn’t change the 43 and 24 pairing as soon as Smid was avaliable but their performance early on in the season looks to have bought them some loyalty.

The PK is awful during this stretch and while the problems are probably systemic, Staios and Horcoff do stand out as being on for a whole lot of chances against.

Friday, January 23, 2009

Scoring Chances - Games 1-10

For those of you reading this blog that are not yet aware, Oilogosphere stalwart, Dennis, has been tracking scoring chances for the Oilers this season and posting them here. It's all very good stuff and well worth checking out. Tracking scoring chances is helpful because it gives observers one more indicator for how a player is performing. Obviously, things like time on ice, the quality of one's linemates and the quality of one's competition will have a big impact on how well any individual player does by this measure, and a useful guide for those measures can be found here. In the coming days I will be posting a tallying of these results in various forms. First up, is the scoring chances in ten game segments.

(For those unfamiliar with the metric a player is awarded a chance any time someone on the ice has a chance to score. He is awarded with a “chance for” if someone on his team has a chance and awarded with a “chance against” if someone on the opposing team has a chance to score. The results are broken down into three game states, EV (even strength), PP (Power Play) and SH (Short-Handed). The players are all organized according to their jersey numbers.)

Games One Through Ten, October 12 to November 1

These first ten games have quite a few Oilers off to a fast start, especially the Kid Line. Although they weren't scoring as often as one would like, all of Cogliano, Nilsson and Gagner have outstanding numbers at even strength. Meanwhile, Horcoff/Hemsky look like they're handling the tougher minutes quite well over this stretch. Pisani is struggling with his new assignment at center, while Penner seems to have adjusted well to playing RW instead of LW. Cole really struggles through this segment and Moreau is off to a poor start.

The D are all off to a pretty good start. Strudwick and Staios run into problems later in the year, but through these first ten games they had been effective at EV (and abominable on the PK, but they aren't the only ones). It helps to make sense of why MacT waited so long to split them up.

Tuesday, January 20, 2009

Oilers v. Blue Jackets - Exodus 20:1-6

And God spoke all these words:

"I am Yahweh your God, who brought you out of the WHA, out of the greedy hands of the EIG. You shall have no other teams before me. You shall not make for yourself an idol in the form of a Blue Jacket or a Coyote or a Shark. You shall not write a blog about them or otherwise follow them at the expense of my chosen people; for I, Yahweh your God, am a jealous God, treating the children of those who forsake me as Flames fans to the third and fourth generation but showing love to a thousand generations of those who love me and keep my commandments."

Sunday, January 18, 2009

Oilers v. Coyotes - Philippians 3:1-3

My fellow Oilogospherians, rejoice in Yahweh! It is no trouble for me to write Biblical Oilerpretation again, and it helps to keep us all faithful. Watch out for those Coyotes, those men who do evil, those mutilators of the legacy of the Great One. For it is we who are chosen, we who play by the Spirit of God, who glory in victory through God's Son, and who put confidence, not in percentages - whether save or shooting - but in God's goodness to his chosen.

Friday, January 16, 2009

Oilers v. Avalanche - Psalm 51:1-7

Have mercy on us, O God,
for the sake of your unfailing love;
yes, for the sake of your great compassion
blot out our 5-1 defeat.

Wash away all of our penalties
and cleanse us from our defensive lapses.

For we know our transgressions,
because our captain is spouting off continually.

Yet against you only have we sinned
and lost to the Wild, which is evil in your sight,
so that you are proved right when you speak
and justified when you judge.

I was sinful in the first half of the season,
sinful, even in the pre-season.

Surely you desire victory for your children;
so teach us wisdom in the defensive zone.

Cleanse our concussions, and we will be better;
wash away our mental mistakes, and we will defeat the Avalanche.

Thursday, January 15, 2009

Oilers v. Wild - 1 Corinthians 9:7-10

Which Oiler skates without knowledge? Which Oiler hits without hope? Which Oiler outshoots the opposition without a reward? Do I say this merely from a human point of view? Doesn't the Torah say the same thing? For it is written in the Law of Moses: "Do not muzzle an ox while it is treading out the grain." Is it about oxen that God is concerned? Surely he says this for us, doesn't he? Yes, and this passage was written for Reddox, because when each Oiler skates against the Wild he ought to do so with the hope of sharing in the victory.

Tuesday, January 13, 2009

Oilers v. Capitals - Deuteronomy 21:22-23

If a man is guilty of a Capital offense and puts another team to death and jumps against the glass, you must not watch the highlights overnight. Instead, be sure to curse him that same day, because anyone guilty of a Capital offense is under God's curse. You must not desecrate the team Yahweh your God has given you as an inheritance by admiring that of another.

Monday, January 12, 2009

Christmas and Coaches

Well that was a wonderful break over Christmas, much of it spent playing knee-hockey with my nephews and Blades of Steel on the old Nintendo. Good times had by all and a nice break from heavy-duty Oiler watching (and work). I probably could have/should have mustered something about the meaning of Christmas, but I ended up just taking a break instead.

Last night's game was very interesting and I think, on the level of the coach, speaks to the experience of the fan. I was over in Lowetide's gameday thread and noticed some positive commentary about Andy Murray, the coach of the St. Louis Blues. Here are a couple of samples:

"I know this will raise some hackles but Andy Murray is a terrific coach. This club is riddled with injuries - McDonald, Kariya, Brewer, McKee, Johnson and now Steen.And they're busting their asses, playing physical, playing well." (Black Dog Hates Skunks)

"Blues fans are clamouring for Murray's head right now; I'm really not sure why. He's a perpetually good coach." (Jonathan)

"That's exactly what I remember about the man, helluva coach and his LA teams at least seemed very motivated and never quit." (Uni)

Now compare this with some of the comments about one Craig MacTavish that have been floating around the internet. The basic sentiment among many Oiler fans is that MacTavish needs to go. He's had a long string of mediocre squads achieve mediocre results and he's been making certain questionable decisions. Now I don't know about the decision-making that comes from Murray, but his career results are nothing magical compared to MacTavish:

Now obviously a coach can only do so much, but Andy Murray's record of mediocrity would seem to be pretty comparable to that of MacTavish.

In the micro view we have last night where Murray's team scored the first goal on the road and then blew it. They still entered the third period in a tie and with hope... but they were outshot in the third period 11-3 and spent most of the time in their own end against a mediocre team missing it's best player. If this is the Oilers on the road in Columbus minus Nash, it's a sign that MacTavish has lost the room and everyone is up in arms. Not to mention, Murray's team is in last place in the conference.

The thing is Andy Murray is a good coach. The best way I know to compare (with thanks to the Contrarian Goaltender) would be to compare results immediately before and after a coach's firing. It would be ideal if I had more examples of Murray's work, but he's only coached in two NHL cities, so the sample is somewhat limited, although on two occasions we have the benefit of him coaching and not coaching a given team in the same season. So here's Murray's first and last year in LA along with his first year in St. Louis compared to LA the year before and after he left along with the year before his arrival in St. Louis:

So it would seem that Murray is getting pretty good results when compared to others (Larry Robinson, Mike Kitchen and John Torchetti) with the same or similar personnel. I guess the rambling point that I'm trying to make here is that just as Murray probably isn't the problem in St. Louis, MacTavish probably isn't the problem in Edmonton. Sometimes the players (like this year's Blues) just aren't very good.