Killing time in the offseason

These are the worst months of anyone's life. Barring Draft Day and the first week of July, there is really nothing hockey fans have to be excited about. It's like a summer-long feeling of boredom similar to having everyone you've ever known out on vacation except you. To pass the time you rerun games in your head or scour Youtube for every hockey highlight reel, even the ones of like 6 year old kids their parents think are amazing.

This lad and his friends took a different approach. Three others and myself have decided that when NHL2009 releases on Sept. 9, we would all make hybrid teams. A hybrid team is a team that is comprised half of our favorite teams (in this case, Buffalo, Washington, St. Louis and Pittsburgh), and half of another team in the NHL that we like. My Buffalo friend chose Columbus, my Washington friend chose Phoenix, my St. Louis friend chose Chicago and I settled on Calgary. Why those three chose their teams is anyone's guess, but I have always had an affinity for Calgary as my Western Conference team of choice. Whether it's because Jarome Iginla is a beast, their color scheme was appealing to my eyes as a child, or deep down inside I am a homosexual and the idea of being a 'Flame' rests easy with me, I cannot say. But I sure hope it's because Jarome Iginla is a beast.

The rules of our little game were simple: 6 forwards from each team, 3 defenseman from each team, and a goalie from each team. Lines were allowed to be comprised however the GM fancied, but the team MUST be salary-cap compliant. And not the NHL08/NHL09 cap, the actual NHL cap hit for players(i.e. Malkin takes a cap hit of 3.834 million, not .984). Also, the most up-to-date roster is in effect.

My team(with ratings), which I created on NHL08 and named the Flint Firebirds:

Cammalleri(88) - Crosby(94) - Iginla(93)
Staal(82) - Malkin(91) - Langkow(86)
Pesonen(82) - Talbot(72) - Fedotenko(80)
Nystrom(69) - Prust(70) - Moss(75)
Gonchar(86) - Phaneuf(85)
Letang(72) - Eriksson(73)
Goligoski(69) - Baldwin(67)
Fleury(84)
McElhinney(71)

My Buffalo friend's team, which I named the Buffalo Generals is as follows:

Nash(85) - Roy(78) - Vanek(84)
Hecht(84) - Peca(86) - Pominville(79)
Huselius(79) - Umberger(75) - Afinogenov(84)
Boll(70) - Murray(69) - KALETA(68)
Commodore(80) - Spacek(86)
Rivet(81) - Tyutin(74)
Klesla(77) - Sekera(71)
Miller(90)
Norrena(78)

My Capital's and Blue's friends are still finalizing their teams and therefore will be posted later.

Obviously, some trash-talking occured over who's team was better. It was in agreement that my team was top-heavy, meaning my top two forward lines were amazing, as well as my top defensive pairing, but that the rest of my team was lacking. The Generals were seen as a deeper, more rounded team with a better defensive core. To solve this dilemma, I decided to undertake a project.

Simply taking the ratings of players and averaging them out doesn't really do the team justice. Therefore, I decided to run some simulations. For each team I did the following. I ran 4 CPU vs. CPU games against the Ottawa Senators on Superstar mode with 5 min periods. After each game, I recorded the TOI in seconds. Two games were played at home, and two were played away. Knowing that there are 3600 seconds in a 60-min hockey game (NHL08's TOI is tallied as if the game were a full 60-min game), I calculated that each second was worth .02777778% of the game's time. Next, I took each player's TOI and multiplied it by that percentage to see their time on ice relative to the maximum 100%. After obtaining this number, I then multiplied that number by the player's skill to evaluate their personal contribution on ice. For example, if a 90 rated player logged 20:00 of ice time in a game, then his number would be 30, since that's how good the team would be if he was the only one who ever went on the ice. After getting each player's "impact" number, I then added up the numbers of all the forwards and divided by 3, since there are 3 men on a line, and divided the defensemen's numbers by 2, 2 men each pairing. The result was a number constituting the average skill of any line/pairing during that game.

This process was repeated for all 4 games and then those final numbers were then averaged. Following will be 3 sets of numbers for the Firebirds and the Generals (Offense, Defense, Total Defense) with 2 numbers each - the average number of their skills, and my numbers("actual"). Total Defense was calculated by adding the Average and Actual numbers of the Defense and Goalie together and averaging.

Fire. - Offense: Average - 81.3333, Actual - 84.9400463
Fire. - Defense: Average - 75.1666, Actual - 77.1836805
Fire. - Total Defense: Average - 79.583333, Actual - 80.59184025

Gen. - Offense: Average - 78.41666, Actual - 80.7318287
Gen. - Defense: Average - 78.16666, Actual - 77.4599653
Gen. - Total Defense: Average - 84.083333, Actual - 83.72998265

As you can see, my offense is marginally better, our defenses are roughly the same, but his total defense is much better. To determine who has the better team, I then took each team's offensive number and subtracted the other team's Total Defense number from it. Whoever has the higher number after this procedure is the better team.

Firebirds - 84.9400463 - 83.72998265 = 1.21006365
Generals - 80.7318287 - 80.59184025 = 0.13998845

As the results show, my team is better by a skill of roughly 1.0700752. This really isn't too significant, but gives me enough to brag. Booyah!

Some suggestions for anyone interested in doing their own project similar to this one:
*For more accuracy, do full-length games and calculate the time spent on the PK. This time should be factored with each players PK TOI and then have the sum divided by 2 since only 2 forwards are on the ice for a PK.
*Run more games for a better balance of stats