Another Look At Quarter-by-Quarter Performance

It's been a while since we looked at teams' quarter-by-quarter performances. This blog looks to redress this deficiency.

(By the way, the Alternative Premierships data is available as a PDF download on the MAFL Stats website .)

The table below includes each teams' percentage by quarter and its win-draw-lose record by quarter as at the end of the 14th round:

(The comments in the right-hand column in some cases make comparisons to a team's performance after Round 7. This was the subject of an earlier blog.)

Geelong, St Kilda and, to a lesser extent, Adelaide, are the kings/queens of the 1st quarter. The Cats and the Saints have both won 11 of 14 first terms, whereas the Crows, despite recording an impressive 133 percentage, have won just 8 of 14, a record that surprisingly has been matched by the 11th-placed Hawks. The Hawks however, when bad have been very, very bad, and so have a 1st quarter percentage of just 89.

Second quarters have been the province of the ladder's top 3 teams. The Saints have the best percentage (176) but the Cats have the best win-draw-lose record (10-1-3). Carlton, though 7th on the ladder, have the 5th best percentage in 2nd quarters and the equal-2nd best win-draw-lose record.

St Kilda have also dominated in the 3rd quarter racking up a league-best percentage of 186 and a 10-0-4 win-draw-lose record. Geelong and Collingwood have also established 10-0-4 records in this quarter. The Lions, though managing only a 9-1-4 win-draw-lose record, have racked up the second-best percentage in the league for this quarter (160).

Final terms, which have been far less important this year than in seasons past, have been most dominated by St Kilda and the Bulldogs in terms of percentage, and by the Dogs and Carlton in terms of win-draw-lose records.

As you'd expect, the poorer teams have tended to do poorly across all terms, though some better-positioned teams have also had troublesome quarters.

For example, amongst those teams in the ladder's top 8 or thereabouts, the Lions, the Dons and Port have all generally failed to start well, recording sub-90 percentages and 50% or worse win-draw-lose performances.

The Dons and Sydney have both struggled in 2nd terms, winning no more than 5 of them and, in the Dons' case, also drawing one.

Adelaide and Port have found 3rd terms most disagreeable, winning only, respectively, 6 and 5 of them and in so doing producing percentages of around 75.

No top-ranked team has truly flopped in the final term, though the Lions' performance is conspicuous because it has resulted in a sub-100 percentage and a 6-0-8 win-draw-lose record.

Finally, in terms of quarters won, Geelong leads on 39 followed by the Saints on 38. There's then a gap back to the Dogs and the Pies on 32.5, and then Carlton, somewhat surprisingly given its ladder position, on 32. Melbourne have only the 3rd worst performance in terms of total quarters won. They're on 19.5, ahead of Richmond on 19 and the Roos on just 16.5. That means, in an average game, the Roos can be expected to win just 1.2 quarters. Eleven of the 16.5 quarters won have come in the first half of games so, to date anyway, Roos supporters could comfortably leave at the main change without much risk of missing a winning Roos quarter or half.

More Green Shoots

This weekend we learned a great deal.

We learned - if we didn't already know - that the Saints are genuine contenders for the flag this year and aren't at all intimidated by the Cats' incredible win-loss record since 2007. We learned that it is possible to make money without relying on occasional improbable victories by rank outsiders and that last weekend's profit wasn't merely a statistical anomaly. We learned that the Hawks aren't good enough to go back-to-back and that the Dees aren't yet focussed on priority draft picks. And we learned that dogs of dubious breeding shouldn't be entrusted with large sums of cash.

Still, it was another glorious weekend's wagering. Combined, across this weekend and last, Investors have landed 31 winners from 33 bets, missing only a line bet on the Cats last weekend and a highly speculative head-to-head bet on the Tigers at $3.75 this weekend. Even then, neither of these bets lost by more than 3 goals.

New Heritage has so far this season won 38 from 49 bets and is up about 45%. Prudence has won 35 from 44 bets and is up about 19%. Hope, inactive this week, has won 7 from 16 bets across the season and is up by about 33%. Line Redux has won 20 from 39 bets and is finally back in the black, although only by about 1%. Chi-squared is what could politely be called the 'outlier'. It's won 7 from 14 bets and is down by around 33%, over one-half of that loss attributable to this weekend's (obvious in hindsight) folly.

For most Investors, the weekend's successes added between about 8c and 9c to their share price. The two exceptions were MIN #002, who was betless this weekend and so marked time, and MIN #017, whose Heritage-only portfolio leapt over 27c.

These increases leave most Investors up by between about 12% and 15% on the season, the exceptions yet again being MIN #002, who's up 33%, and MIN #017, who's up nearly 45%.

Here's a chart showing each Investor's triumphal march:



On tipping, this week's best performance was HSH's 7 from 8. Most other top tipsters scored 6 from 8, though BKB managed just 5, dropping it into joint 3rd place with STM II on 78 from 112 (70%). Shadow and Silhouette remain joint leaders, now on 82 from 112 (73%). CTL fills out the top 5 places with 76 from 112 (68%).

Only one tipster, ELO, is now showing a loss on level-stake, home team only wagering, commencing in Round 6. Applying this same wagering strategy to the tips of our top 3 tipsters would have yielded profits of over 11 units, which represents an ROI of over 10%.

Whilst ELO might not be performing well on head-to-head wagering, it continues to excel on line betting, recording 6 from 8 this weekend to move to 65 and 47 for the season and a profit to level-staking of 10.57 units. Chi had a good line-betting week too, picking 7 from 8 this week, but he's still at less than 50% (he's 54 and 58) for the season.

In running down the Dockers in the final term on Sunday night, Carlton became only the 8th team this season to come from behind at three-quarter time to win and the 1st team to do so when trailing by 2 goals or more at the last change.

As I've commented before, this has not been a season for winning after trailing. About 68% of teams who've led at the first change (ignoring games that were tied at this change) have gone on to win. This is about 1% point higher than the all-time average. Teams leading at the half (again ignoring games tied at the half) have won 81% of the time, up 2% points on the all-time average, and teams leading at the final change (ignoring games tied at this point) have won 93% of the time, up a startling 7% points on the all-time average.

This latter statistic is not, as you might suspect, the result of teams tending to lead by more at the final change this season compared to previous seasons. In fact, even teams with quite slender leads have been hanging on to win at historically high rates. For example, teams leading by between 6 and 11 points have won 85% of the time (up 13% points on the all-time average), and those leading by between 12 and 17 points have won 100% of the time (up 18% points on the all-time average).

The only three-quarter time lead range that has proved more vulnerable this season than has been the case historically is a lead of between 1 and 5 points, which has been defended only 55% of the time this season, 3% points under the all-time average.

AFL Players Don't Shave

In a famous - some might say, infamous - paper by Wolfers he analysed the results of 44,120 NCAA Division I basketball games on which public betting was possible, looking for signs of "point shaving".

Point shaving occurs when a favoured team plays well enough to win, but deliberately not quite well enough to cover the spread. In his first paragraph he states: "Initial evidence suggests that point shaving may be quite widespread". Unsurprisingly, such a conclusion created considerable alarm and led, amongst a slew of furious rebuttals, to a paper by sabermetrician Phil Birnbaum refuting Wolfers' claim. This, in turn, led to a counter-rebuttal by Wolfers.

Wolfers' claim is based on a simple finding: in the games that he looked at, strong favourites - which he defines as those giving more than 12 points start - narrowly fail to cover the spread significantly more often than they narrowly cover the spread. The "significance" of the difference is in a statistical sense and relies on the assumption that the handicap-adjusted victory margin for favourites has a zero mean, normal distribution.

He excludes narrow favourites from his analysis on the basis that, since they give relatively little start, there's too great a risk that an attempt at point-shaving will cascade into a loss not just on handicap but outright. Point-shavers, he contends, are happy to facilitate a loss on handicap but not at the risk of missing out on the competition points altogether and of heightening the levels of suspicion about the outcome generally.

I have collected over three-and-a-half seasons of TAB Sporsbet handicapping data and results, so I thought I'd perform a Wolfers style analysis on it. From the outset I should note that one major drawback of performing this analysis on the AFL is that there are multiple line markets on AFL games and they regularly offer different points start. So, any conclusions we draw will be relevant only in the context of the starts offered by TAB Sportsbet. A "narrow shaving" if you will.

In adapting Wolfers' approach to AFL I have defined a "strong favourite" as a team giving more than 2 goals start though, from a point-shaving perspective, the conclusion is the same if we define it more restrictively. Also, I've defined "narrow victory" with respect to the handicap as one by less than 6 points. With these definitions, the key numbers in the table below are those in the box shaded grey.

These numbers tell us that there have been 27(13+4+10) games in which the favourite has given 12.5 points or more start and has won, by has won narrowly by enough to cover the spread. As well, there have been 24(11+7+6) games in which the favourite has given 12.5 points or more start and has won, but has narrowly not won by enough to cover the spread. In this admittedly small sample of just 51 games, there is then no statistical evidence at all of any point-shaving going on. In truth if there was any such behaviour occurring it would need to be near-endemic to show up in a sample this small lest it be washed out by the underlying variability.

So, no smoking gun there - not even a faint whiff of gunpowder ...

The table does, however, offer one intriguing insight, albeit that it only whispers it.

The final column contains the percentage of the time that favourites have managed to cover the spread for the given range of handicaps. So, for example, favourites giving 6.5 points start have covered the spread 53% of the time. Bear in mind that these percentages should be about 50%, give or take some statistically variability, lest they be financially exploitable.

It's the next percentage down that's the tantalising one. Favourites giving 7.5 to 11.5 points start have, over the period 2006 to Round 13 of 2009, covered the spread only 41% of the time. That percentage is statistically significantly different from 50% at roughly the 5% level (using a two-tailed test in case you were wondering). If this failure to cover continues at this rate into the future, that's a seriously exploitable discrepancy.

To check if what we've found is merely a single-year phenomenon, let's take a look at the year-by-year data. In 2006, 7.5-to 11.5-point favourites covered on only 12 of 35 occasions (34%). In 2007, they covered in 17 of 38 (45%), while in 2008 they covered in 12 of 28 (43%). This year, to date they've covered in 6 of 15 (40%). So there's a thread of consistency there. Worth keeping an eye on, I'd say.

Another striking feature of this final column is how the percentage of time that the favourites cover tends to increase with the size of the start offered and only crosses 50% for the uppermost category, suggesting perhaps a reticence on the part of TAB Sportsbet to offer appropriately large starts for very strong favourites. Note though that the discrepancy for the 24.5 points or more category is not statistically significant.

Chi Returns As Hope Fades

Chi, as a dog (which he is, so I guess that's a little redundant, but bear with me) is prone to inexplicable bursts of canine euphoria during which times he's apt to bound around like a dog a fraction of his age.

Whatever it is that transports him into these displays of friskiness clearly struck sometime this week just as he was contemplating his weekly wagers. Why else would he have decided to put 17.7% of the Fund on the Tigers at $3.75? His other wager is a much more sedate one: 4.6% of the Fund on the Pies at $1.55.

Prudence has sipped from the same chalice - dog slobber and all - as evidenced by its two most extreme wagers of 1.9% on St Kilda at $2.35, and 1.5% on the Dees at $2.10. As well as these it has three other wagers on teams at prices ranging from $1.30 to $1.55, the largest on being 4.9% on the Swans at $1.30.

New Heritage has also found a lot to like, with wagers in all but the Richmond v Adelaide game. These wagers include almost 12% of the Fund risked on teams priced above $2: Melbourne (5.4% @ $2.10), Port (3.1% @ $2.35) and St Kilda (3.1% @ $2.35). The four other bets range in size from 11.4% on Sydney @ $1.30 to a neither-here-nor-there 0.1% on Carlton @ $1.55.

Line Redux, the only other active Fund this weekend as we are again Hope-less, has found four wagers, three on teams receiving start and one on the Pies, giving start.

Together these produce a Ready Reckoner that looks a lot like this:



Unusually, due mainly to Chi's Richmond wager, all Investors except MIN #017 have roughly equal upside and downside risk this weekend.

On tipping we have:
* Collingwood the unanimous favourites over the Dons.
* West Coast 9-4 favourites over the Dees. Amongst the top six tipsters, the Dees have Shadow, Silhouette and STM II as supporters and the Eagles have BKB, ELO and CTL. Chi has this as one of his four - yes, four - one point margin predictions and hence Games of the Round.
* Brisbane are 12-1 favourites over Port. Only HSH is tipping Port.
* Adelaide are 12-1 favourites over the Tigers. Chi is the sole Tigers fan (hence the wager) but here too is prediction a single point margin making this another of his Games of the Round.
* The Dogs and unanimous favourites over the Hawks.
* Sydney are 8-5 favourites over the Roos. The top six tipsters are split as for the Eagles v Dees clash. This time Shadow, Silhouette and STM II are siding with the Roos and BKB, ELO and CTL are on the Swans.
* St Kilda are 10-3 favourites over the Cats. Geelong have, amongst the top six tipsters, only BKB and ELO in their corner. Chi and ELO are tipping just one point margins and so have this as a/the Game of the Round.
* Carlton are 11-2 favourites over Freo. Only HSH and EI II have tipped Fremantle. Chi though has this as another one pointer, which makes this his fourth and final Game of the Round.

So close are many of the games expected to be this weekend, anyone who tips the card deserves double points in their tipping competition. One measure of this expected closeness is the average points start per game on line betting.This week that statistic is 14.6 points per game, more than two points lower than the next-lowest this season, which was 16.9 points per game in Round 5.

Another measure is the combined probability of all underdogs winning in the round. The probability of that happening this week is about 3,800 to 1, which is roughly 12 surprisal bits. Previously, the lowest this has been this season was about 7,950 to 1, or approximately 13 surprisal bits, which it was in Round 1. In other words, the probability of all 8 underdogs winning this week is roughly twice what it's been for any other round this season.

For Line Betting:
* Chi's on Collingwood, Melbourne, Port, Richmond, the Dogs, the Roos, the Saints and Fremantle.
* ELO agrees except for swapping West Coast in for Melbourne.

Almost Perfect

Wagering weekends don't come much sweeter than this last one. Only Geelong's lack of killer instinct - which surely must be troubling the Cats' coaches and fans alike - spoiled a 15 wins from 15 wagers result.

Still, 14 from 15 was good enough to allow all Investors to record solid gains for the week in amounts ranging from about 5% to 8.5%. On the back of these gains the Recommended Portfolio is now up by over 4% and all other Portfolios are up by between 6% and 33%.

At the Fund level, Hope rose by the largest percentage, climbing 8.5% on the strength of its sole wager on the Dons. The Dons have now contributed more than 34c worth of gains to the Hope Fund, the result of 3 wins from 5 wagers. Overall, the Hope Fund now has a 7 and 9 record.

The next highest increase was recorded by Prudence, whose 7% increase this week thanks to 6 successful wagers, was its greatest single-week increase this season. Prudence now has a 5 from 5 record with St Kilda, a 3 from 3 record with Geelong, and a 30 from 39 record overall.

New Heritage bagged 5 from 5 and jumped 6.3%. It now has perfect records with Geelong (3 from 3), St Kilda (5 from 5), Collingwood (3 from 3), and the Lions (3 from 3). Its season-long record stands at 31 from 42.

The only other active Fund this weekend was the Line Redux Fund, which managed 2 wins from 3 thus registering only its third profitable round in the last 10. It now stands at 16 from 35 for the season.

On tipping, Silhouette and Shadow each scored 7 from 8, moving them to 76 from 104 (73%) for the season, one further tip clear of BKB, whose 6 from 8 left it on 73 from 104 (70%). STM II also scored 7 from 8 and is now on 72 from 104 (69%). Only one tipster picked all 8 winners this weekend: HSH, who now sits on 60 from 104 (58%). No tipster fared worse than 6 from 8.

All that fine tipping has lifted the wagering returns from level-staking our tipsters' tips. Nine tipsters are now showing a profit on level-stake, home-team only wagering, commencing in Round 6, and four of those are also showing a profit on level-stake, season-long wagering.

Moving to line betting, Chi and ELO each tipped 5 from 8 this weekend, which leaves ELO at 59 from 104 for the season and Chi at 47 from 104. Level-stake wagering on ELO line bets is now showing a +7.17 unit return for the season.

Lastly, an update on the Mean Absolute Prediction Error metric. Chi's MAPE is currently 30.9 points per game while ELO's is an exceptional 27.9 points per game, only narrowly trailing BKB's 27.7 points per game.

No Triskaidekaphobia Here

Clearly the Funds aren't superstitious. Why else would they, collectively, choose Round 13 as the round in which to - for most Investors - make their largest outlay?

Foremost for profligacy amongst the Funds is the New Heritage Fund, which this week has found five wagers totalling approximately 64% of Initial Funds at prices ranging from $1.04 (for the Cats) to $1.22 (for Adelaide). So, mercifully, not much in terms of risk for each bet, but a considerable amount in terms of total risk. Indeed, this is the largest amount that the Heritage Fund has wagered in any single round this season.

Prudence, too, has set a record for most Initial Funds at risk this week. It's wagered about 31% of the Fund on six wagers, five of them on the same teams as the New Heritage Fund, but the sixth on West Coast at - wait for it - $3. Since when does Prudence wager on 2/1 longshots?

The Hope Fund is a little more restrained, making just a single wager on Essendon at $2.65 for a smidgeon over 5% of the Fund.

Also somewhat restrained is the Line Redux Fund with 3 wagers this week totalling 15% of the Fund. Two of these wagers are on short-priced favourites offering 44.5 (Brisbane) and 47.5 (Geelong) points, and the third is on the Roos receiving 36.5 points.

Coyest of all is the Chi-squared Fund, which has found nothing to wager on this week.

As I hinted at earlier, combined these wagers represent the largest single-round outlay for all Investors except MIN #002. The table below provides detail about the outlay and return for each Investor profile for each round.

On then to this week's Ready Reckoner:

So, for those with the Recommended Portfolio and for MIN #001 and MIN #015, the Geelong v Port clash is that contest representing the largest potential swing between the best and worst results. For this same group of Investors, the Brisbane v Melbourne matchup represents the next largest swing.

(This week the Ready Reckoner includes information about the best- and worst-possible outcomes for each Investor across the weekend. For example, a perfect set of eight results for those with the Recommended Portfolio would lead to about a 7% gain and, by contrast, a anti-perfect set of eight result would lead to just over a 24% drop.)

To tipping:

* Essendon, the market underdogs, are 9-4 favourites over the Blues. Amongst our best tipsters, BKB, ELO and CTL are all siding with the Blues, while Silhouette, Shadow and STM II are with the Dons. ELO, though coming down in favour of the Blues, has them winning by just 2 points, making this its Game of the Round.

* Collingwood are unanimous favourites over Freo.

* Adelaide are unanimous favourites over the Swans.

* Brisbane, continuing the trend, are unanimous favourites over the Dees.

* Hawthorn are 11-2 favourites over the Eagles. Our two least-performed tipsters in HSH and EI I are the Eagles' only friends.

* Geelong, as is custom, are unanimous favourites over Port.

* The Dogs are the round's fifth unanimous favourites, they over the Roos. Chi, however, has the Dogs winning by just 4 points and therefore has this as his Game of the Round.

* St Kilda are 12-1 favourites over Richmond, the Tiger's sole supporter being EI I, which this year is just about as close to no support at all as it's possible to have.

The only game with any significant implications for our top tipsters is, therefore, Friday nights Dons v Blues clash.

On Line Betting this week, Chi's with all the teams receiving start. ELO differs only in that it's predicting the joint competition leaders in the Cats and the Saints to cover the spread.

Swans Turn Into Ugly Ducklings

Two of the three teams did what Investors asked of them this weekend, but Sydney's failure to be amongst this group meant another small loss for most this weekend. This leaves the Recommended Portfolio down about 1% on the season and all other Investors up by amounts ranging from 0.9% (MIN #015) to 24.6% (MIN #002).

At this point in the season New Heritage is up about 11%, Prudence is up about 3.5%, Hope is up almost 25%, Chi-squared is down about 17.5%, and the Line Redux Fund is down about 21%.

On tipping, Silhouette and Shadow continue to perform well and jointly lead our tipping competition, two tips clear of BKB and four tips clear of STM II, who've leaped into fourth thanks to an 8 from 8 performance this round. Over the past 3 rounds STM II has scored 21 from 24.

On level-stake, home-team-only wagering, commencing in Round 6, we still have five tipsters showing a profit: Shadow, Silhouette, BKB, STM II and EI II. Three of these tipsters - Silhouette, Shadow and STM II - are also showing a profit on season-long wagering.

ELO performed well again this round on line betting, picking up 6 from 8 to go to 54 and 42 for the season, which is good enough for a 5.67 unit profit on level-stake wagering across the season. Chi managed just 4 from 8 and is now 42 and 54 for the season.