Dear Diary Picks Its Poison

Submitted by Seth on May 22nd, 2011 at 7:53 PM

theerieottersbus existentialism

I don't know and don't want to know why it turned up pounds of shirtless man meat on page 1, but add "OHL" to things you should never search for on Google Images. "Existentialism" on the other hand, is quite entertaining.

Dear Diary,

Say you're a top Ontario hockey prospect. Your dream is to play for Michigan, but Michigan is not even allowed to contact you:

Stacked somewhere in between the Oshawa Generals and Bowling Green was a letter with a block M in the corner. It contained a brochure for the University, a questionnaire, and the contact information for the coaches. Turns out, I had not been emailing coach Berenson at all, but now I had his real address. I emailed him and got a response from an assistant coach. He told me to call him.

WHAT?! I can call these people?! How was this not explained to me before? I had never bothered to email any other school and since the player must initiate contact, they couldn't reach out to me.

Silk-Wolves-Sign-475You  don't know if Michigan will even offer you because it's too early on NCAA's schedule for anything like certainty. You are surrounded by people who want to see you in the OHL. You are drafted, and given a contract:

I sat down with the GM, who knew that I was considering college. He basically explained to me the benefits of the CHL, the education packages, and the unique experience of being a young local celebrity. He was very polite about it, but told me that if I was signing, he wanted it done within 2 weeks.

The diaries were blessed again with the concluding Parts II and III of JimLahey's epic personal tale that illustrates just how difficult it really is for hockey talent to cross the border, even for those who desperately want to play NCAA. Well written and poignant, with everything from twists to cliffhangers to a surprise ending, its plot is worthy of an episode of Law & Order. Except it happened to this guy. This guy: Diarist of the Week (again).

There was only one other diary this week, but it's a good 'un: ebv returned to do another analysis of correlation between defensive talent and performance, and also defensive experience and performance. Ganking charts:

Experience / Talent:

He used the Rivals depth charts so that's guys on the two-deep, not the starters. The comments had a lot of suggestions, and this overlaps with them, but I think there are a couple of factors that really need analysis more than average age and star-rating of two-deep:

  • Size of classes over 5-year period. A 3-star who made the two-deep out of 108 potential defensive players recruited is probably going to be more qualified than a 3-star who made the two-deep out of 40 defensive players recruited.
  • Run it again with just starters. The two-deep still includes a lot of guys who might not be ready to play, because the scholly limit and dress limit and whatnot. If the great teams have a 5-star junior or senior starting and two hyped freshmen backing them up, it won't show with an average age.
  • Gimme a "worst starter" breakdown too. Xcalibur once tried to test that when Michigan was rolling out Kovacs at free safety and every team we faced began running most of its plays directly at Kovacs. If there's a "weak link" effect for defenses, that will throw off your performance metrics for the team defense.
  • Related to above: Distribution?
      5th yr Sr Sr / RS Jr Jr / RS So So / RS Fr True Fr Avg.
    Team A 3 5 5 5 4 2.91
    Team B 9 1 1 1 10 2.91

    You are looking at teams with very big differences. Team B can field a much more experienced starting 11, but any injury or bust or low-rated older guy means a true freshman is likely starting. Team A's distribution is far more likely to be seeing players in their middle years handling significant time. You can do the same thing with star ratings:

      5-star 4-star 3-star 2-star Walk-On Avg.
    Team A 3 5 5 5 4 2.91
    Team B 9 1 1 1 10 2.91

    Any busts from B and you are starting a 2-star or a Walk-On. Distribution among the starters would tell us if there's a weak link effect too.

After the break (in honor of jg2112's poor scrolling wheel) I realign the NHL and suggest a playoff system.

OT: Realigning the NHL - This Is My Thing


Last week the board got into a bit of an NHL realignment discussion on a thread wondering if the Red Wings might have any shot of ever getting back to the East. It's been a pet project of mine to put together a new NHL alignment and playoff system. I was going to mention it in the thread, but I realized I've never really put down my plan on a blog before. A few Red Wings blogs have offered the forum, but since I'm staff here I figured why let someone else have it, right? If the NHL to you is just a place for former Michigan hockey players to continue their careers, skip to the playoff part – the same concept might work for any post-season. Like in football too.

My concept: 0 conferences, 5 divisions of 6 teams, and a pick-your-poison playoff draft system that allows top seeds to choose their opponents.



Norris (aka "Erie-Ontario"): OTT, TOR, PIT, DET, CMB, BUF

Patrick-Adams (aka "Atlantic"): NJ, NYR, NYI, MON, BOS, PHI

Heartland: MIN, CHI, DAL, COL, WPG, PHX

Gretzky (aka "Pacific"): VAN, CAL, EDM, ANA, SJ, LA

This supposes Atlanta moves to Winnipeg, which is an announced unannounced thing. If that doesn't happen, swap out Atlanta/Winnipeg for St. Louis. The divisions don't really matter as much as the concept (an Original Six Division is possible).

The reasoning for going to six-team divisions is two-fold: it makes scheduling easier, and it reduces the effect of divisional imbalance. It's still possible to have the six worst teams in one division, or the six best in another, but it's less likely to get a matching six than a matching five in one place.


  • Home and home with every team in the league every year (58 games)
  • 4 additional games versus division rivals (20 games)
  • One extra home and home series with rotating pool of ex-conference rivals (4 games)

Equals 82 games. (Option: in Olympic years the extra two series can be dropped to make it 78 games).

The last is a little complicated but think of it as a vestige of the old conferences. It rotates against your old conference foes (because it sets up reunions from old playoffs), just 2 a year so it doesn't significantly change the schedule. Each team has 14 other ex-conference members so the rotation will re-set after every 7 years. This is the only time an East or West designation is counted for teams.

One year for Detroit: Six games each versus Ottawa, Toronto, Pittsburgh, Columbus and Buffalo, four games versus Colorado and St. Louis, and two games versus every other team in the NHL.

Playoffs (this is my favorite part):

Since there are no conferences, the playoff matchups are determined from a Pick Your Poison Draft.

Who Makes the Playoffs? The five conference winners plus 11 at-large bids as determined by:

  • Most points
  • Most wins
  • Most goals for
  • Head-to-Head
  • Toughest schedule (avg. pts of opponents)

How Seeds Are Determined: The President's Trophy Winner gets to select the team they want to play from among the 5th best division winner and the 11 at-large teams (based on most points, current tiebreakers from entire NHL).

Then the next best division champ picks the team they want to play in the 1st round. Then the next, and so on. After the 4th team makes their selection it goes to the teams with the highest point totals (so if you're the worst division winner you basically just got yourself an at-large spot).

Try it with this year's standings. Seeds:

Seed Team Division W L OTL Pts GF
1 Vancouver* Gretzky 54 19 9 117 262
2 Washington* Dixie 48 23 11 107 224
3 Pittsburgh* Norris 49 25 8 106 238
4 Philadelphia* Atlantic 47 23 12 106 259
5 San Jose Gretzky 48 25 9 105 248
6 Detroit Norris 47 25 10 104 261
7 Tampa Bay Dixie 46 25 11 103 247
8 Boston Atlantic 46 25 11 103 246
9 Anaheim Gretzky 47 30 5 99 239
10 Nashville Dixie 44 27 11 99 219
11 Phoenix* Heartland 43 26 13 99 231
12 Los Angeles Gretzky 46 30 6 98 219
13 Chicago Heartland 44 29 9 97 258
14 Montreal Atlantic 44 30 8 96 216
15 Buffalo Norris 43 29 10 96 245
16 Dallas Heartland 42 29 11 95 227

I put stars next to the division champs. It didn't matter since they were the top point-getters anyway, but I gave priority to the Top Four division champs to preserve the meaning of winning a division championship. It's 4 not 5 because I want to penalize that one crappy division champion we get every year. Some years a team who won an easier division will get to pick ahead, but what I like about it is there's incentive for a team cruising to the top of an easy division to keep trying to wrack up points.

Also different between this and current: Dallas (95 pts) is in, the Rangers (93 pts) are out.

At the end of each round there's a re-draft. Teams retain their seeds through the entire playoffs. The selecting (higher-seeded) team gets the extra home game.

Why This?:

The Pick Your Poison approach has several advantages over the current conference system:

  • Teams are likely to select rivals and nearby cities. Proximity and rivalry increase fan interest and ticket sales, and lower travel costs across the league.
  • The regular season means more, because a playoff seed means more. Every team finishes the season jockeying for some type of position (best record overall, win division, be the better division winner, get to a "selection" seed, etc.
  • Teams would no longer "lose" their way to a better playoff matchup, as teams have done in the past.
  • Top seeds can avoid the "trap" series, like that 8-seed who's been unbeatable since February, or the 7-seed known for injuring your star players (COUGH Anaheim).
  • More teams per division and fewer divisions means less variation between divisions, thus fewer seasons when entire divisions are playoff-bound, or division champs with fewer points than the lowest at-large team.
  • Fuel for the fire in every series: kind of hard to say you respect your opponent when you specifically chose them as the team you thought easiest to defeat. The selectors come in arrogant, the selectees with a chip on their shoulders. Pissed off players = good hockey.
  • It makes the race for the President's Trophy a bigger deal because you're not just ensuring home ice, but getting to manipulate the playoff picture.


What if a high seed picks another high seed? This is the biggest downside. Say in our 2010-'11 seeding above that Washington selects Pittsburgh, thus effectively making Pittsburgh's 3-seeded season less valuable than Buffalo's 15th. It would be a rare team that wants a hard matchup to start the playoffs but it's theoretically possible. The tradeoff is that would be an awesome series for the NHL, creating a first round matchup with high national interest.

More importantly, the pick-your-poison system is tough on teams that put up better records overall, but who come into the playoffs injured. Imagine a two-seed whose best player won't be available until the 2nd round, e.g. if Washington was missing Ovechkin. Vancouver might choose to knock them off right away, but the season Washington had should have earned them a relatively easy first round matchup. There's a simple fix: Guarantee Seeds 2-4 immDSC01183unity from 1st round selection. The 5th best team is still able to be nicked but at least they had a chance to avoid it had they won their division.

What about teams who choose matchups for non-competitive reasons? So say Columbus has a dream season, gets a 4-seed, and selects 5th seed Detroit because a series against the Red Wings would guarantee more playoff revenue? This sucks for Detroit especially, as the NHL team most likely to be selected out of its earned seed by a small market team looking to cash in. Again, Detroit can avoid this with a higher seed. And Columbus is taking an awful gamble that they'll be knocked out of the playoffs, meaning Detroit effectively can move up a seed by winning.

Also I want this to be as feasible as possible, and that means all of the NHL's needs must be served. When you're pitching this to Bettman, it is very important to point out that the new system still accomplishes his goal of screwing the Red Wings.

Jokes and tinfoil hats aside the biggest knock on my system is that how likely a team is to win is not the same as how many games a team won. For the top seeds this is made up for and more, because winning more during the season will translate into more control over your playoff destiny. For the lower seeds, their seasons mean less. I think it's a fair tradeoff. The 5th seed is vulnerable, but still in a much better position than the 15th seed. Ultimately do you care more that a 7th seed gets a worse matchup than a 13th seed, or that the top seed gets a better matchup than the 4th seed?

This is About Pro Hockey, Why is it on a Michigan Blog?

Because NCAA could do this too. It wouldn't be nearly as accurate or work as well because the system for choosing seeds in NCAA is a hell of a lot harder than an organized league with a draft where all teams play each other. Since these are 1- and done games there's no incentive for the high seeds to select teams that are nearby. However it does remove some of the burden of figuring out seeding on the lower end, and puts it on the shoulders of the high seeds. I imagine they'd be more interested in selecting the devil you know, so this means more rivalry games early in the tournament. And by giving top seeds a bigger advantage you'll see more of them in the Frozen Four.

The concept also might be applied to a college football playoff. Take 12 teams by BCS rating with the best four Division Champs given a bye week. The next four pick their 1st round nemeses, and play at home. In Round 2 the following week the top four draft their opponents and play at home (again). Round 3 is another draft, and takes place at the Rose, Sugar, Orange, and Fiesta, and then the champ game's at a rotating site. I kind of like it for college football because the difference between the BCS No. 6 and No. 8 team in season performance is probably so miniscule the best way to tell who's better is to simply let another team make that call.




May 22nd, 2011 at 8:25 PM ^

...especially when you only had it as a 5-day rental from VideoMax and your parents insisted most of that time be used for going to school, doing homework, cleaning the room, etc. We once got as far as aquiring the hammer before the game was stopped because 6-year-old sister wanted to play Bubble Bobble.


May 23rd, 2011 at 10:05 AM ^

I like the playoff idea, but there is must be a reason that matchups in playoffs are always set and not chosen by the higher-seeded team. I think allowing the teams to choose who they play, draft-style, can harm the integrity of the playoff system because teams can choose the team they play for reasons outside of competitiveness. Owners can pick teams like the Wings that draw well so they increase their ticket sales, or try to play teams that are close to reduce costs. Having an opponent draft opens up the possibility of thinking the league is fixed, or subject to smoke-filled room scenarios. Gambling is another big issues the leagues would rather not even come close to dealing with.

As for the divisions, I like the 2 conference 3 division idea, but I am going to try to do it a bit differently. I want to basically make it North/South, or the Canadian teams and no Canadian teams conferences. 


Northwest division stays the same (VAN, EDM, CAL, MIN, COL)

Northeast stays the same (BOS, BUF, TOR, OTT, MTL)

North-Central CHI, DET, NYR, NYI, NJ


Southwest same (SJ, LA, ANA, PHX, DAL

Southeast is the same (FLA, TB, WAS, CAR, ATL [note: I thought of this before Atlanta started making moves to Winnipeg. Swap them with NJ if they go to WPG])

South-central CBJ, NSH, PIT, PHI, STL

Obviously the North is stacked, but that's kind of the point. It increases the number of games between the Canadian teams in the East and West (four between each, instead of only 1 or two now). Plus, the conference keeps the Original Six together. As for the South, it would probably be tough for them to survive without having the draw of the Wings, Maple Leafs and Canadiens coming each year, but a more balanced schedule could help. 

The North/South conference alignment helps balance the travel imbalance that hurts the West teams, especially in the playoffs. It's not a perfect solution, but I think it's better than what we have now.


May 23rd, 2011 at 12:06 PM ^

Okay I'm a hockey fanatic. I watched every Wings game this year (thank you Sling Box + DVR) and I would love for the Wings to be in the East (7pm starts not 10:30pm) and we'd get to play Pittsburgh, Washington, Boston, Toronto, Montreal, etc but your playoff scheme is crazy.

Hockey does need a new playoff system. Home field advantage has vanished and all it takes is a hot goalie to completely devalue 82 regular season games. I think everyone loves underdogs but at the same time having a #1 seed should mean something. Sometimes the #8 seed is more feared than a #7 seed or even a #5 seed. Often the #4 seed is a better team than the #3 seed.  That shouldn't be the case. The fact that each division winner has a locked top 3 is also a major problem. Often the #3 seed would finish in the bottom half of that conference but because their conference sucks they get a top 3 lock. 

I think hockey should look to the Scottish Premier League for an answer to their regular season problems. Each team plays ever team in their division 6 times and plays the rest of their conference 3 times. After 54 games (6*4 division teams +3*10 conference teams) the league splits into two. The top 7 of the East and the top 7 of the West would then wait while the bottom 16 teams would play a single elimination playoff for the final 15th promotional spot. Then the top 15 teams would play each other twice as would the bottom 15 (for 28 more games, 54 + 28 = 82 regular season games). After the 82 games we'd have a regular season champion that should mean something considering 35% of the games would be the best playing the best. 

Then the top 15 teams + the winner from the bottom 15 would play a 16 team mixed conference playoff. This would be similar to the current Stanley Cup finals except there would only be one bracket. The teams would be ranked depending on their standing from the 28 games only. So a regular season champion might not be ranked first. I don't know if re-rankings should occur, or if the bracket should stay locked like in NCAA Basketball or if the rank should stay with the winning team (so no matter what the #1 seed is the winner from the 1/16 game no matter if the #1 or the #16 team wins). 

I think this would make both the regular season mean something and still have an awesome playoffs.

Just an idea I concocted with a friend over drinks.