2010 NBA Playoffs Predictions: Conference Semifinals

posted in NBA

With the first round now over, we’ve updated our 2010 NBA playoffs predictions for the eight remaining teams.

First, let’s recap the first round and check in on Henry Abbott’s TrueHoop Stat Geek Smackdown 2010 competition over on ESPN.com. We’re happy to report that we correctly picked the winner of all eight first round series. In addition, we correctly picked the length of four of the series.

Those results are good enough to land us in first place in the TrueHoop Stat Geek Smackdown. The scoring system for the smackdown is 5 points for each correct pick plus a 2 point bonus for picking the correct length of the series. Our total for the first round is 48 points – putting us 4 points ahead of John Hollinger (who currently leads the “official” contest).

As we look to the Conference Seminfinals, here are our picks:

– Cleveland/Boston: Cleveland in 6
– LA Lakers/Utah: Utah in 7
– Phoenix/San Antonio: Phoenix in 7 (barely … very close to Phoenix in 6)
– Orlando/Atlanta: Orlando in 6

As we mentioned in our previous post, Utah is most likely the contrarian pick. Our algorithms see the series as being razor thin close, but ultimately gives Utah the slight edge. Two further comments:

(1) Our algorithms do not make any specific adjustments for injuries – both Utah and LA have some injuries they are working through, although Utah has taken the bigger hit.

(2) We did some further analysis that indicates if the series does indeed go 7 games, the Lakers would have the edge in that deciding game because it would be played on their home floor.

With all of that said, we stick by the numbers here at TeamRankings.com and will stick with the Utah in 7 pick.

Remember to keep tabs on how series win odds change after every completed game with our NBA Playoffs Current Series Win Odds.

  • Hex

    Guarantee Utah won\’t win. Orlando is going to be Atlanta in 4, maybe 5.

  • Hex

    Utah is going to be lucky to win 1 game in this series.

  • Cleveland_Fan

    Utah in 7 has got to be the stupidest prediction ever. Who ever picks the lakers to lose is just a hater and i bet if there was money on the line you guys would get your heads out of your asses and pick lakers in 5, 6 maybe.

  • http://www.teamrankings.com matt

    Hex / Cleveland_Fan – thank you for your comments and thoughts.

    We realize the Utah in 7 pick is controversial, but we certainly don’t think its the “stupidest prediction ever”. Also, we make our picks strictly be the numbers (for better or for worse – often for better!) – in no way is any Laker hatred factored in (even if two of us did grow up in the Boston area!)

    Bottomline: we see this as a close series – and we’re not alone. If you check out the other stat geeks in the TrueHoop Stat Geek Smackdown, most have this series going 7 games. (See here; http://sports.espn.go.com/nba/playoffs/2010/news/story?page=Smackdown-10 and select the Lakers/Jazz series at the bottom).

    As we mentioned in our blog post, one weakness is our algorithms is that we don’t make any adjustments for injuries other than how the teams have been playing over the past few weeks. Injuries could certainly play a factor in this series.

    The only other caveat that we’d add is that upon further analysis we feel that one of the follow scenarios is most likely:

    (1) Utah in 6.

    (2) Lakers in 7.

    Due to home court advantage if the series actually does go to 7 games, the Lakers probably have the advantage. With that said, we don’t change our picks after we’ve made them, so we are sticking with the Utah in 7.

    We look forward to seeing what happens – if Game 1 is any indication, it should be an entertaining series…

  • Cleveland_Fan

    Yeah, seems like a close series right?!!

  • cleveland_fan

    yeah you’re right, the lakers are up 3-0 but the jazz could still win in 7 [edited by TeamRankings]

  • http://www.teamrankings.com matt

    Cleveland_fan – we don’t mind the critique of our picks, but please keep it clean.

    Despite the Lakers being up 3-0, I think anyone watching the games would say that the series has been close. (The 111-110 and the Lakers needing an 8-2 run in the final minute to win Game 3 is just one concrete example).

    Our algorithms said the teams were pretty evenly matched (again with the caveat for injuries) and we think watching the series play out has shown that. Yes, the Lakers have won each game, but the games have been close overall. (This is a much closer series, for example, then what we are seeing with Orlando and Atlanta).

    With that said, at the end of the day, the games are played and there is a winner and a loser. Would the Lakers be up 3-0 every time if we replayed this series 100 times? I think you would agree that the answer is most likely no.

    And, in fact, that’s exactly how we made our pick. We simulated the series 10,000 times. I guarantee you that in some of those simulations the Lakers were up 3-0. However, the majority of those simulations had the series lasting 6 or 7 games – and by a very slim margin our power ratings gave Utah an advantage. The power rating difference was well within our margin of error, but at the end of the day we make picks and our system is to make that pick based on the numbers. And so we did.

    We’ll leave the final post mortem until the series is actually finished and see where we stand. In the meantime, enjoy the games…

  • Cleveland_fan

    haha, your still trying to back up your ignorance about the playoff. haha. I didnt know that this website was for entertainment purposes. Thank you for the laugh guys.

  • http://www.teamrankings.com matt

    We were wondering when you were going to stop by for your victory lap.

    Yes, we were wrong about the outcome of the Lakers/Utah series. We stand by the explanations we gave above (as well as the caveats in that particular pick that we explained since we initially made it). No need to rehash, at the end of the day the models didn’t have all of the information they needed to correctly assess the series.

    With that said we are still 10-1 in correctly picking the winners in the 2010 NBA Playoffs (and we’ve picked 4 of those series to exact number of games correct).

    Beyond that, if you are interested in more than just entertainment, our Decision Tree model has been doing great throughout the playoffs. Over the past 21 days:

    20-11-1 (64.5%) against the spread
    30-17 (63.8%) against the totals line
    39-8 (83.0%) in straight up game winners

    (Link: http://www.teamrankings.com/nba/model/59/overview)

    Stay tuned for what we have to say about the Lakers/Suns series…