Cann Tables

As regular readers will know by now, I am not a big fan of the conventional League Tables. For betting purposes, they are at best of little value, and at worst, downright misleading. I won't mention the futility of using them before the season has already started, but late in the season, their use is also limited. The biggest problems I see with them is that the recent form is hidden, and the difference between teams is hard to identify. The tables are listed by teams with their corresponding points totals, and teams separated by goal difference look exactly the same as teams separated by twenty points. The excellent A Beautiful Numbers Game today had a post on Cann Tables, which addresses this problem by listing the points totals for teams, and then the teams (if any) with that total. An example is this one from Ian Eliowart's website:

This does show at a glance how much, for example, the Liverpool clubs are adrift of the top six - something not immediately obvious from the conventional league table. My own preference is to used a modified Elo ratings system, which has its disadvantages (e.g. a sudden dip in results can be slow to be reflected if the underlying performance remains strong, or a dramatic change in team personnel will not be immediately reflected, but overall it works well).

They didn't work so well today though, with the run of Unders on the XX Draw selections coming to an end in the Bundesliga at Schalke '04 v Borussia Dortmund (1-2). It is quite likely that the Bundesliga will be dropped at the end of the season. With a strike rate of 18 from 63 after today, they are the weakest of the five leagues I follow, hardly surprising when they had an average goals per game of 2.87 heading in to today's games. Ian Erskine selected the 3-3 draw between Bayer Leverkusen and Hertha Berlin as his lay selection, a result that not only bumps up the average GPG in Germany, but also increases the profits from backing these lay selections!

I have been trialling the inclusion of more matches in the non-Bundesliga leagues. Prior to today, the results over three seasons of the Classic XX Selections would have been 110 winners from 334 selections (32.9% - 3.04). Using the Expanded Selection criteria from England, France, Italy and Spain, the results would have been 301 winners from 926 selections (32.5% - 3.08). Similar strike rate, but a lot more selections, an important consideration. It is also worth noting that the average price on Bundesliga selections is correspondingly higher than in other leagues:

If I continue the service next season, it will likely be modified along these lines. I'll be e-mailing subscribers for their thoughts and ideas before making a decision. I've been most impressed with the lack of complaints of whining when results are poor, something I was concerned might happen when the inevitable losing runs at these prices come along.

I had some more selections in Italy this weekend, but following the death of Livorno's Piermario Morosini today, it looks as if these games will be postponed.

Another long-priced (33/1) winner today in the Grand National, of academic interest only to yours truly, but another ding against anyone suggesting that this race is anything but entertainment and a lottery. Al took my post title of National Lottery a little too literally:

Cass old boy, I would say that betting on the national is nothing like doing the national lottery.

For one, you have a 44/1 chance of picking a winner and 7m/1 chance of winning the lotto.

The odds of winning the lotto are so stacked against you that there is no value in even playing unless the jackpot goes up to a certain level.

Not so much of the 'old boy' thank you. I am getting quite sensitive about my age these days. My comment that

The chances of finding a winner may not technically be in the 'lottery' realm, but as sports go, they're not far off.

was very much tongue-in-cheek, although the inability of many to differentiate between say a 1 in a 1,000 and a 1 in 10,000 shot is well known.