ESPN Playoffs
Today I was looking at ESPN and I noticed they had a piece called the Expert Picks in the playoffs section. So I checked it out to see which teams these so called "experts" picked to advance from the first round.
It was pretty sad. I mean, I think all of our picks are better and we don't watch every game let alone write articles for ESPN. There are 10 experts on the page. I'll go over a few of their "awesome" picks:
7/10 experts picked the Suns to beat the Spurs. Good call fellas. Idiots.
6/10 experts picked Dallas to beat New Orleans. Dallas? Really? Isn't Dallas the team that we were HOPING the Rockets could play in the first round? Hasn't Dallas totally gotten worse since the Kidd trade? Weren't they like 0-10 against the other West teams in the playoffs? Didn't they get beat in the first round last year by Golden State? How the hell could you pick Dallas to beat NO? Especially 6 out of 10 so called Experts.
5/10 experts picked the Wizards to win. This one is not as bad and I know the Wiz could still win, so we'll see. But I'm pretty sure Cleveland has this series in the bag.
6/10 experts picked Detroit to win but in 5 games, which can't happen now. I know what you're saying, "At least they picked Detroit." But how much did we talk about how good Philly was before the playoffs started? Patrick had Detroit in 7 and Jeremy picked them in 6.
The last one I'd like to point out is that all 10 experts picked the Jazz to win. Which, obviously is wrong. Just you wait and see!
Subscribe to:
Post Comments (Atom)
9 comments:
HAHA! Word up! Houston all the way baby!
Haha, jimi thinks i am stupid:) I had a feel of those picks.
No wait, i had all of those picks. Though for the first three, i was kinda 50/50 either way, and picked them to go 7.
Jimi, you picked Detroit in 5 too. Welcome to the club.
I'm gonna post everyone's first round predictions after its over. Should be interesting.
Well I was going to try to make a mini game of it but I'm having issues with doing the 2nd round percentages right now. If I reformat my original table, it may help.
Anyhow, our original predictions are still posted here:
http://spreadsheets.google.com/pub?key=pMP-TZSjHcRJZFWywjZa7XQ
and the 3rd tab now has a point value listed for each matchup in round 1 and I've totaled it too.
I weighted the series winner at 50% and the number of games played at 50% (so each game is worth 1/7th of a point * 50%). For this reason, any series that isn't complete will show a pretty low percent (as nobody has been awarded the .5 for a correct prediction) - if you really want to know how your doing on those, just multiply by two for now and that will give you your games accuracy at least.
Nice to know 4 out of 5 fans can out predict the expert [Hollinger] though!
If I have enough time, maybe I can import he ESPN analysts too to see who did the best overall.
Well... I went ahead and added a 4th tab labeled 'experts' and it looks like Jimi's exactly correct. We're better than the 'experts' and Jimi's better than all of'em!
I did notice a couple of things, I guess googledocs doesn't have some of the functions that excel does (SUMPRODUCT) so it messed up the 2nd tab a little. It also doesn't like merged cells so some of the formatting didn't transfer. Otherwise, it seems to be pretty decent.
Jeremy! Explain that spreadsheet
_1st tab_
Simply a visual representation of our predictions (laid out like a mini tourney card) with the total number of predicted games in the row of the predicted winner and the column of the round. This is just for record, I didn't do any calculations on this tab. (last row and column were just sanity checks to verify that I carried the appropriate number of games to the other tabs when I broke our predictions apart)
_2nd tab_
Our predictions with a 1 (for an entire series win) in the row of the predicted winner and the column of the round. This is probably the most basic but the function didn’t make the transition to googledocs. Maybe after the entire season is over, I’ll upload a data only file so it’ll show better.
_3rd & 4th tabs_
This tab was originally intended to predict the success of various teams based on our opinions (hence the name ‘Success Predictors’). You know, the whole 5 brains is better than 1 theory? I'm pretty sure there’s good evidence that says the size of a sampling is directly proportional to the odds groups have of predicting things.
The columns all the way to the right reflect 'strength' based on a 7-game series equaling 1 total point (.3 points for the series win & .1 for each game). The total for each round is multiplied by the round number (so predictions of a round 3 win are worth 3 times more than a round 1 win). Intuitively, I know neither step was the best statistical choice but it was only meant to be a quick representation. The rounds are totaled and then averaged (divided by 6 including Hollinger). I went ahead and put a % of max (being if everyone picked them to win it all).
After the first round, I figured there’d be a good way to show how accurate our predictions were. I created an “Actuals” column to compare things to. It got real complex real quick because of the way I displayed the data to begin with. It wouldn’t have been that bad if we all predicted the same teams to win but because reality is fluid, the cells that I compare to each other had to be auto adjusting. It worked for the 1st round but after that it got stupid. I know how to get it done but the formula becomes so long that after I've created it once it's really difficult for me to duplicate to the rest of the series because it's so hard to parse. If I rearrange the data I can get it done. I’ll probably give it a go after round 2.
Either way, round one is done. Nailing the prediction gives you 100% accuracy. Getting the winner correct but the # games wrong will subtract a small percentage. I weighted the series winner at 50% and the number of games at 50%. This really give the series winner prediction 50% + (4/7)*50% because you would have at least 4 of the 7 games correct. The best example of this in the first round is Boston. Most people predicted 4-0 to Bos. Everyone got the .5 (or 50% of a point) for the series winner. The other half of the point is a % of correct games. So everyone got 4 of the 7 games correct, or: (4/7)*.5 [or .285]. Added to the 50% for the winner, the total is 78.5% (or .785). Jake wisely predicted ATL to take a game so he got 5/7 for the games giving him an extra 1/14th of a point (or 1/7th * .5) added giving him 85.7% (or .857). Get the series wrong and the most you can get is 6/14th of a point (or .429) and that’s only if it goes 7 games and you picked 7 games to begin with (just the wrong winner of game 7).
Probably not the greatest explanation but hopefully you get it. If not, feel free to ask again and I’ll try to clarify.
Post a Comment