2008 SCORING

One of the features of the current pick 'em scoring system (let's call it option A), with the points per win doubling each round, is that entrants can make huge moves up and down the table during the later rounds and, to a large extent, you are always in with a chance of making a significant move up the table until your last surviving pick goes out.

That's great, since it tends to keep most people interested all the way through, but it can sometimes be a bit brutal. It doesn't mean that whether or not you pick the correct champion is all that matters (a lot of pick 'ems turn out to have been decided by R1 or R2 scores if lots of people pick the same champion), but getting the correct champion is hugely important, perhaps too important?

Probably the most accurate measure of prediction skill is how many picks people get right, as shown in the 'Correct Picks' column, but because a large majority of the matches in a tournament are played in the first two rounds, I think basing the final results on one point per pick (let's call that option B) would kill the competition for most of the entrants too early on.

If we are going to change the scoring system at all, the end of the first year is reasonably sensible point at which to do it, so we have been looking at possible 'compromises' that might give us the best of both worlds.

Options we discounted

Basing points per pick on the Race points for reaching each round (e.g. 7, 15, 25, 45, 70, 100 for an AMS) would be a bit less extreme than the existing system, but probably not enough to make it worth the complication of changing.

Basing points per pick on the change in race points for reaching each round (i.e. 7, 8, 10, 20, 25, 30 for an AMS) would be even less extreme, but the ATP's increments are a bit illogical, e.g. 7 -> 8, then later on a big leap from 10 -> 20, then back to 20 -> 25 and we think this would seem too complicated as well.

Our preferred option (at the moment)

We think the best scoring system might be a simple 1 for a R1 win, 2 for a R2 win, 3 for a R3 win, 4 for a R4 win and so on, let's call that Option C. It still makes picking the correct winner and the correct finalists important, but not in quite such an extreme way, and still ensures some big moves up and down the table late in the tournament, but again without being quite so extreme.

Example - Cincy 2007

As an example, have a look at the Cincy pick 'em results. In this pick 'em, all of those who had picked Fed to win ended up in the top 35 and all of those who picked anyone else to win ended up below all of Fed-to-win entrants. The latter group included Rochelle, who was leading all the way up to the Final but experienced a brutal drop from 1st to 36th through not having picked Fed to win, ending up below a couple of entrants who got just 20 picks correct throughout the tournament compared to Rochelle's 30!

Changing from the 1 - 2 - 4 - 8 - 16 - 32 to the 1 - 2 - 3 - 4 - 5 - 6 system would have produced these alternative results (the change in position column shows the effect of the change in scoring system in this case) - as you can see, the top half of the table is still dominated by Feddies, but Rochelle would only have dropped three places to 4th on the final day and at least a couple more of those who got the champion pick wrong would have ended up in the top half, with positions more evenly spread between those with different winner picks in the bottom half. Have a look and see if you think it looks fairer or not.

N.B. Please try to look at this without paying too much attention to whether you or your friends would have done better or worse under the new system yourself - those who would have done better under the new system at Cincy could well do worse under it in other comps and vice versa!

Just to be absolutely clear, if we do change the scoring system, it will be for next year, we will NOT recalculate past tournaments like this and we will NOT change the names of past winners! As in tennis, we play to the rules as they are at the time!

Comments please!

Anyway, comments are very welcome before we make a final decision, including suggestions we haven't thought of yet. We might even put it to a vote on the AO pick 'em entry form if it's clear that there is no consensus.

P.S. Anyone worried that any change might mean a lot of work, thank you for thinking of that, but don't worry - a change to the scoring system (as long as it is as simple as options B or C) is one of the easiest things to do on the spreadsheet and takes a matter of seconds.

P.P.S. To anyone perceptive enough to have already worked out that this change will probably make all the final scores closer and hence usually reduce the bonus ranking points a future tournament winner would get, thus disadvantaging them against those who won tournaments this year while they still have their points on the system, that can easily be coped with by increasing the number of bonus points awarded per 0.1% above the average for next year after analysing the differences it would have made this year, so it's not a problem.