It’s weird being our first summer on Twitter and not having those regular European fixtures everyday to keep us busy – at this rate I’ll start reflecting on my life. As our promo showed, our global coverage means we still track plenty of leagues inplay but loads are over in South America kicking off through the night so not the most convenient, i think that’s why we get ultra critical feedback in the replies if any lose / goal too early!
You may have read a few posts ago I wrote about tracking the wins and losses based on the BetScore %. I think by the end of the European season the sample is becoming reasonable to hypothesis around what we’re seeing.
The main focus of our service is of course over 0.5, so that’s how I’ve measured success of our BetScore’s tipped. The below graph shows the average winning and losing BetScore % tip since start of 2017:
Seems on over 0.5, we do have some surprising results when high pressure occurs in the 1st half of games. As you can see, the data points look a bit like a heart rate on the flight home from a stag do but there are some lines of best fit there. The difference in them shows that the average losing BetScore % is actually higher than the winning. Maybe this is where the favourite in these games causes the lesser team to shut up shop soon after half time or if the game is more even, then maybe it just dies out as so much effort is spent going all guns blazing in the 1st half.
When looking at the BetScore’s in intervals of 10% and how they perform, there does seem a sweet spot of the 60-70% range where the win rate is 85%. That’s over 5 wins for each loss and it will be interesting to see if it carries on to next season. We have a balance to find here at Panel HQ where we want to adjust the BetScore algorithm where we see improvements but also need to wait long enough to see a true reflection of what’s going on.
A caveat to this discussion though is although the sample size is larger at this point, the last half of it was through the relegation scraps at season-end. We saw plenty of weird results through that period with more factors in play as some teams had nothing to play for and some with everything to play for. Those with everything seemed to be more likely to defend early and stay resolute to rack up every point possible. It will be interesting to see if we have some irregularities at season-start as well as teams find their feet.
It’s also important to note that this graph does not discuss success of over 1.5. A potential next area I’ll look into is how the BetScore % would compare for winners and losers in this market including the success spread across score intervals. As if you think about how a game pans out, where the big gaps in quality and therefore pressure exist, there is usually a greater likilihood in either staying nil nil if the poor teams parks the bus, or having more than one goal once the deadlock is broken. So there could even be some value on backing the horrible unders market if there is a big gulf of in-play stats, especially if there has been a red card. We won’t go auto-tipping that just yet but we don’t see these factors being considered by the bookies.
Try to enjoy the internationals this weekend and hopefully we’ll have more action from the server in the daytime!