So, if you're a follower of Punting Form on facebook, you've
probably been seeing the tips coming from the Tipdroid every
morning. And if you've been using those tips to inform your bets,
then you're probably pretty angry by now. And you should be. Angry
at yourself. I mean, I warned you that I was still working on how
to best use the robot and as it turns out, the best way to use it
is to bet against whatever it tells you.
Things started out promising. At first, I was betting on
everything, even the mildest hits, and I was doing pretty well out
of it. I was placing lots of little bets every day and hitting on
enough of them that my betting bank was steadily climbing. But the
thing was, there were just too many runners to have them go up on
facebook and I had to find a way to narrow it down to just the very
best bets. That's where I started to run into trouble because it
turns out that the Tipdroid's built-in methods of measuring
confidence were completely unreliable.
I've been tweaking and testing (and occasionally kicking,
punching and spitting) for a month now and the so-called
"high-confidence tips" that were being pushed through to the
facebook people, well they've just been getting worse and worse.
And as I fiddled with the settings and connections, I managed to
screw up the overall performance to the point where my dozens of
little bets weren't doing so well anymore. Now I could just reset
the thing to how it was before and I'm sure I'd do just fine again,
but...apparently it's not all about me and I'm supposed to be
providing a service here. So what to do?
The answer, I think, is to embrace the crappiness of the tips
we've been getting. In fact, I'm going to try and make them even
crappier. The Tipdroid was doing a great job of finding horses that
should win but just don't, so let's run with that. From now on,
we're going to bet against the droid (or at least, avoid betting ON
it). The Tipdroid is dead. Long live the Laydroid.