Gemma's Game (she didn't hang around did she...)



Not a lot of nuance with this banker, and the games are already becoming a bit samey sadly... just takes a pretty strong run of luck to even get to 5 figures now, whereas she'd have probably been offered the governor on the old board, and we'd have had some actual tension. There's nothing special about this version, nothing's on the edge, partly through lack of Edmonds and also the budget money tree.

Anyway, the banker throws money at the end of the game, and can't really blame the player for taking this, a rare draw in this version!

Mike's Game:



Nice comeback to make it a mildly exciting finish from what seemed another trainwreck. Banker loses all concept of maths at 2-box again, and for once the player is sensible enough to take it!

Deal-Eye leads 15-4 with 5 draws.

Orla's Game:



Missing the entire "Power 5" in the opening round and the offer is *still* a grand, consolation at 5-box but most players would overreach for the big sum. It's inevitable we compare to the original version, where the first offer may have been £15k+ and the 5-box offer maybe £10k, with some sort of utility decision to make. There's barely any dilemmas for the contestant with these offers, they're just bad or inconsequential when a board has already crashed. Definitely a shadow of its former self.

I'll come back to this if there's any insightful analysis to be had...

Current score: 16-4 with 5 draws

BRING BACK DEAL-EYE!

Well... quite a bit has happened since Orla's Game, much of which has been unfavourable for Deal-Eye, and it has been largely powerless against.

With the banker often overspending on the box from about 5-box onwards (sometimes earlier), and some of the players being ultra-aggressive to the end (complete opposite of the end of the Edmonds era!), and the £250k no longer the jackpot, it's unable to grab its massive victories of old.

After 50 shows, Deal-Eye has 22 wins, the Player has 16 and there have been 12 draws (1 win and loss from Celeb shows)

Deal-Eye has won £472,652.00, at an average of £9,453.04
Players have won £539,642.11, at an average of £10,383.50

This is as much an indication of poor Banker strategy than anything else, particular a recent game featuring a £37,450 offer on a £50/£75k - we've seen offers as low as £15k on that finish in the past!

Generally, Deal-Eye has pedantically beaten the player or accepted low stakes on a weak board most human players would risk utility-wise. There's not the nuance to the game anymore, the banker seems to hope for an early collapse or blindly throw ITV's coffers at the game for a feel-good result at the end. Neither adopting the mantra of "buying the box for as little money as possible"...

There's no way you can legitimately claim Deal-Eye is useless on this version though. There's been a lot of variety as to when it deals between 8 to 2 box and a pretty even amount of wins and losses, plus it's fun comparing the human approach to the analytical approach at these slightly lower stakes.

At the very least it certainly beats late 2013-early 2014 where it went to the end in every game.

Yes, I suppose you could say Deal-Eye gave an exhibition of its abilities that day! Any human statisticians here must feel denuded by its computerised genius; their failings exposed ruthlessly, whether they consented to it or not!

All that matters to Deal-Eye is that it has its way. If it had feelings, perhaps it would get a real kick out of showing off its prime assets to unsuspecting people - teasing, teasing again, and finally working up to the big climax for all to see...
It's interesting because, while its crown jewels - in this case, the algorithms that make so many of its predictions so accurate - are undeniably beautiful, Deal-Eye itself (particularly on the face of it) is nauseatingly ugly to look at! A kind of Frankenstein's Monster made real, if you will...

But the amazing abilities of Deal-Eye to judge when is the right time to deal or not can't be questioned. I just wish we had something like Deal-Eye to help us decide which humans are safe to trust, and which ones we should run as far away from, as quickly as we can.

Don't you agree, Andy?