I had been wondering about a graphical way to show how streaky the preds are since the post by AUarchitect. Finally came up with this implementation of the up/down ticks. This is current through the win over Colorado last night and the small bumps are OT losses. Downloaded stats from Hockey reference. Created a new column of values: 1 for win, 0 for OT losses, and -1 for losses. Used the sparkTable package in the statistics program R to create the HTML graphic which was then sent to my flickr account so that it could be inserted here. Far more than you ever wanted to know; I realize, but had make that 75 word limit!

Predators W/L 2010-2011

When your power play ranks 28th out of 30, your team better be scoring some goals during the 5 on 5. Yet the Predators seem to be a team where the ability to put goals in the net can be elusive at times. I was interested in seeing who was contributing during non-power play time given their amount of time on ice. I downloaded stats from NHL.com and divided even strength goals by the players' average per game even strength time on ice. The results show a few surprises. Generally, D-men have lower rates compared to forwards which is expected. However, Shea Weber and Cody Franson rank higher than Ward, Legwand, or Erat.  While the EKG line is clicking, Erat still seems to be under-performning compared to his line mates (though his PP production has been up recently). JP Dumont reportedly  in the coaches' doghouse is in the second tier of scorers. Most surprising to me is that Colin Wilson leads even strength scoring when you factor in his time on ice. And then there is Joel Ward who is apparently suffering from Zidlicky syndrome (get a hefty raise, play like crap).

Note that Spaling, Sulzer, Belak, Lombardi, Begin, Laasko,and Klasen have not scored any even strength goals this season and therefore, are not included.

But you'll miss it in January

Much like the current hot and humid weather in Middle Tennessee, everybody gripes about the Predators' power play  but nobody does anything about it.  With a goal scored on only 15.7 % of power play opportunities, the Predators ended the regular season ranked 24th (1). More dramatically was the fall in the ability of the penalty kill . For several years it was considered a strength of the team, but in 09-10 the PK was successful 77.1% of the time (average for all teams = 81.8%, range=74.6-86.8%). Fortunately, team discipline paid off as the team was the second least penalized team in the league.  Yet as dismal as it seemed last year, the power play was better than the 2 previous seasons. Here are the PP rankings and rates since the lockout.

I didn't recall the team being so futile for so many years until I started looking into these stats. But before you start screaming for Poile to beg Paul Kariya to come back and run the power play, why don't we look at what difference a PP on steroids might have done for the season just passed.

Stamkos to the Predators! (E0)

No not really. One of the things I like about On The Forecheck is that Dirk doesn't dabble in rumors  (now I'm just sucking up). I bring up young Steven Stamkos not because he will soon be enjoying evenings on Broadway with Taylor Swift and looking for a condo in a not yet foreclosed building in the Gulch, but because he scored 24(!) power play goals last season. The most scored by a Predator was 10 by Hornqist (2). Now the Tampa Bay Lightning had 40 more PP opportunities than the Preds but that is a lot of PP goals by one player. The best team performance on the PP was by the Capitals. They had a league-leading conversion rate of 25.2% on the PP. Lets look at this past year and see what would have happened if the Predators had such a power play. Could they have won the division or at least achieved home ice advantage for the first round?

Review and Simulation 

The Predators record was 47-29-6. How many extra wins could have been achieved with a better PP. I reviewed each loss and win in OT/SO to see what the performance was for the PP in those games. My idea was to run a simple simulation of the unsuccessful PP opportunities with a new turbocharged PP which performed at the Capitals level. I examined the results from 47 games. Of those, 18 games were losses by a margin greater than the number of failed PPs or were OT/SO wins over Eastern Conference teams or teams which finished lower than the Preds in the West (extra point given to the losers didn't affect Preds standing). In the 29 remaining games, the Preds PP was awful scoring 7 PP goals in 97 opportunities (7.2%). I next generated random numbers (between 0 and 1)  for each of the 90 unsuccessful power plays and counted a goal scored if the random number was <= to the Caps .25 rate. This resulted in an additional 23 goals being scored (a reasonable rate of .256). However, when added to the 7 goals scored in the actual games, the rate raises .3. To get the rate back down to the advertised .25, I randomly dropped 5 goals to give an overall success rate of .257 for all the games being simulated. These additional goals were enough to change the winner in 5 games and force a tie in an additional 4 games thereby changing the results in 9 of the 29 games. The Preds did remarkably well in OT/SO games winning 70% of them. To resolve the ties I generated 6 more random numbers and counted wins when the random number was <= .7. This resulted in 2 wins and 2 overtime losses. Just as important were the points lost by the opposing teams. Detroit lost 6, Phoenix 3, San Jose 2 and Chicago 2.  While the Detroit numbers may look questionable, most of the games against the Red Wings were very close and the Preds had many PP opportunities from which they did not benefit (1/15 in the 4 games included in the simulation). 

Final Standings and Thoughts

Updating the points changed by the theoretically improved power play, the Preds would have won an additional 7 games and the Western Conference. The simulation I performed is very unsophisticated and run only on the raw numbers looking back at the 09-10 season. It does not take into account the ebb and flow of the game. A good team might respond to being tied or giving up the lead due to an opponent's power play goal with renewed energy and zeal. A home crowd might will the team to victory during a tv timeout (I've seen it happen!). A mascot may throw a cream pie into the face of an unsuspecting fan of the opposing team. All these things can shift the momentum in a game and running such a dry simulation fails to account for many effects which may negate a solitary additional power play goal. I only ran the simulation once due to time constraints. Averaging the results after running it many more times might cause the effect to dampen a bit.  A more sophisticated approach might have adjusted the PP rate based on a given opponent's penalty kill. There are lots of variables one might include or discard and knowing what to do with them is part of the challenge in trying to make statistical sense of such a fast-paced and fluid game.

I don't anticipate the Predators scoring on the PP at a rate comparable to the Caps anytime soon. I have to believe it will be a coaching priority this year. We have seen over the last few years that the Preds are not a high scoring team. Even a modest improvement in the PP could result in a significant increase in points and seating in the post-season. However, as the Caps proved, having the best power play in hockey may not be enough to get you out of the first round.