Sunday, 2 August 2015

NBA 2017 MVP race – part 1

(Originally posted on Thursday, 10 November 2016)

Part 1 can be called “at least 1 loss by every team”. Precisely, it means that the statistical data comes from the day when the last team(s) without a loss recorded their first loss of the season.

All the data I used I found on this site:

I described my way of calculating the MVP value in this post:
Comparing NBA MVPs (and Steve Nash).

It turned out to be a very good way of determining official NBA MVPs – in the last 40 NBA seasons (in the seasons that I had analysed) my calculations were spot-on for 28 seasons. That’s 70 % accuracy! If I included also seasons when the official MVP was second on my list then my calculations were correct for 34 seasons. That’s 85 % accuracy!

The problem is that previously I analysed only whole (finished) seasons and now we are just at the beginning of the season. My formula for MVP value is based on a reference point of 60 wins by the team of a particular player (because players from weak teams have only minimal chance of winning the MVP award and a player from the worst team will never win the MVP award), so I couldn’t use it directly (the current values would be laughably small).

To make the current values comparable to the whole-season values I made mathematical predictions of the future values – I simply assumed that every player will play exactly the same way he’s been playing so far (that he will have the same average statistics for the rest of the season) AND that every team will play exactly the same way it’s been playing so far (that its win percentage will stay [almost exactly] the same for the rest of the season [with round number of wins]).

On 8 November 2016 the Cleveland Cavaliers recorded their first loss of the season and now there are no teams without a loss.

With the assumptions described above, according to the data from the games up to 8 November 2016 the current NBA 2017 MVP race list is this:
1. Russell Westbrook: 3003.11     (projected team wins: 70)
2. LeBron James: 2619.63     (projected team wins: 70)
3. Chris Paul: 2530.52     (projected team wins: 70)
4. Kevin Durant: 2212.35     (projected team wins: 59)
5. James Harden: 2153.84     (projected team wins: 47)
6. Blake Griffin: 2110.41     (projected team wins: 70)
7. Kevin Love: 2098.11     (projected team wins: 70)
8. Kawhi Leonard: 2090.02     (projected team wins: 59)
9. Kemba Walker: 2000.24     (projected team wins: 68)
10. Damian Lillard: 1906.12     (projected team wins: 51)

Russell Westbrook had a monster start of the season by I doubt if he can keep it up for all of the season AND if the Oklahoma City Thunder will have such a high win percentage at the end of the season. LeBron James, Chris Paul and Kevin Durant are strong MVP candidates too, mainly because we can safely assume that their teams will end with lots of wins. If the Houston Rockets improve their win percentage James Harden will be a much stronger MVP candidate than he is right now.

Please notice that there is no Anthony Davis on the list, even though he had a monster start of the season too. The problem is that his team did not win even 1 game, so his current MVP value is 0! Assuming that the New Orleans Pelicans would win 50 games in the season (pretty unrealistic assumption) his MVP value would be only 2252.27. A very high number of team wins is usually decisive and I'm afraid that Anthony Davis will not be a true MVP candidate in the near future.

No comments:

Post a Comment