As I mentioned in the Deal-Eye thread, I would like to see analysis done on the older games of the series and not just those games which aired since Deal-Eye was implemented in March 2011. After an interesting "discussion" where pretty much every flavor of person from respectful to outright mean had a say, I stepped forward as the one to decide to do this retro analysis.

To be clear, this isn't my system or my spreadsheet (though I did make a nice enough facsimile before I received StatsMan's version...). This is all due to, and largely thanks to, StatsMan. Though it can be argued that anyone could have come up with this, he's the one who did, and so I am choosing to associate this to him in every way imaginable. In case this isn't clear enough, thanks StatsMan for creating a wonderful spreadsheet that nicely presents the strength (or lack thereof) of the banker's offers!

My current plan is to roll out a week of old games each weekend. That way, there's no confusion with the current Deal-Eye thread, and there is something to look at during the off days of DoND. This might of course change depending on various factors (most of which will probably be external).

With that said, let us start with...

Lynn (31 October 2005)



We will be seeing a pattern of early banker stinginess leading to late banker generosity. Perhaps this is due to the producers wanting people not to deal the opening offer, since there was no precedent for gameplay at the time. As such, Lynn ends up with a thermo-stat of 0.96 despite the first two offers having respective benchmark percentages of 28% and 19%!

Sadly Deal-Eye loses out at the fourth offer due to the highly volatile board. In fact, it's surprising Lynn did as well as she did considering she only brought £10 to the table...

More games coming later today!

As StatsMan would say, £2,900 (at 8 box) is an offer Deal-eye would be more likely to take than a human, at least before greeny Summer. Personally, I would've gone at £4,300 as it was a very generous offer for the board and I would want to ensure that the first game did not end in a blue win that set precedent for future cautiousness.

Mark (1 November 2005)



Horrible start for the system, yes, but fantastic start for the players with two consecutive banker spankings!

The stingy-->generous offer trend is seen to be really strange here. A £900 opening offer is increased to £1,500 after taking out the £250k, which is then increased again to £6,000 after taking out the £75k (the highest amount in play, since the £100k went in the opening). Then we have another uplift to £8,500 after taking out the second highest amount, which finally was dropped to £5,000 after the highest amount went and it was a £20k-high game.

Actually Mark's story is similar to Lynn's--horrible box selections and a blue in the box, and yet plays to 2-box and deals the highest offer for a spanking. If these games were slightly modified, perhaps there would be no Deal or No Deal beyond November...? (Or perhaps I'm exaggerating the effect of these games. )

Anita's game coming later tonight!

There was always going to be DoND after November 2005 anyway....

Would've been a definite £8,500 for me as well in Mark's game, as over 75% of the average is always a Deal for me unless the board is extremely stable. 5 box, with guaranteed £500 and the highly generous final offers to look forward to back then, would've fallen under that category, so had I gone on I would've dealt the £9,900 like Mark did.

The early days I predict will be notoriously streaky for the system, a lot more clear-cut decisions, and the players could rely on generosity continuing a lot more than nowadays...

So, there's likely to be a lot of early deals where Deal-Eye can't factor this in. I'll just say that Deal-Eye was launched in the first place with a consideration of banker behaviour at the time. I may well have looked at things differently in 2005.

However, it will be interesting to see how long each "wave of wins" lasts, and they'll be very few draws!

I agree completely. If the American Syndicated version was still around, for example, I would probably adjust the formula so that it calculates the greater of the Fair Deal or 80% of the average for each finish, as the latter is around what the American Banker would offer on average on volatile finishes (excluding the obvious hypothetical lies in prove-out ). In fact, I made a few spreadsheets to see how this Deal-Eye would react with American offer patterns and most of the time it would take either one of the first two offers (which no human player was willing to take, of course) or the final offer.

I do wonder how Deal-Eye would've fared if it discarded decisions on generosity of the 1st 2 offers, like virtually every contestant has... but clearly that's not the point of the system!

Any chance you could record these results onto Excel Saperogo, much like I've done for the last 750-odd games..?

Anita (2 November 2005)



The system's first draw, and the players' third consecutive banker spanking to start the series! Though it was a close call at 8-box with the offer coming within £381.38 of the benchmark...

We'll continue with Rachel's game tomorrow!


I can, but what kind of compilation do you want? Just the amounts Deal-Eye won and the players won? Or a complete record of the board at each stage, the offers at each point, etc.?

Yes! The initial commission of 66 shows takes it into 2006!

Yep, a sensible Deal by both the player and Deal-eye in that game. I personally would've chased the 40% chance of a six figure final offer with 30% comfort cushion of a £14,000ish final offer back then but I definitely would Deal £33,000 on that board today. It then would've been a Swappy £3,000 for me in that game and absolutely no regrets.

Just a point about the scoring of draws - I decided not to just because then it's easier to see the number of victories Deal-Eye/ Player have gained overall (ok, I've had 2 0.5 point scenarios for 'contentious' results...). Then you can see the difference between game number and scores as the number of draws. Totally up to you of course, though!

I think 0.5 is good for a draw, that way when you add up the points, you get the number of shows!

Hmmm. I didn't fully realize draws weren't tallied. I remember seeing 0.5 on both sides for a while and thinking that had something to do with a draw.

I think I'll stick with what I have for now, though. The nice thing is I modified the "After xx games" cell to reflect the number of games played, which for now is #(player wins) + #(Deal-Eye wins). (I even went slightly fancier for no reason at all and threw in an IF statement so it says "After 1 game" instead of "After 1 games". ). If we decide later that maybe having a separate entry for #(draws) is preferred, I'll modify as needed. Though I am more fond of giving both sides half points since I think that's what's done in other competitions?

All I thought is that it gives you EXTRA information about the make-up of points that the other method doesn't - I know the Ryder Cup scores every game, but in a sense that's a competition with an overall points goal. This isn't, so I preferred to highlight victories.

If you kept a record, you could always state the number of draws periodically anyway!

But it is a competition with an overall points goal...?

Rachel (3 November 2005)



Deal-Eye opening its box for once and winning the first red brought to the table, £20,000. Unfortunately, Rachel beat it by £5,000 with a fifth-offer deal. Compared with yesterday's £33,000 offer with 2 blues/£3k/£35k/£250k, the £25,000 offer here is surprisingly stingy. Regardless, the only way to win more in this game was to deal at £40,000, which is difficult with a £20,000 guarantee.

Tally after 4 games
0 wins
3 losses
1 draw
Deal-Eye: £64,400
Players: £81,900

Well... Essentially, it runs 'infinitely', or at least a pre-determined target/ goal hasn't been set! Any side can theoretically overtake at different junctures of the experiment. And if you can monitor wins over particular eras, it can only be a useful thing I think.

Definitely would've been a Swappy £75,000 for me in Rachel's game. Her 5 box offer was very stingy, and I certainly would not want to be the person who showed the Banker that he could get away with stingy offers (like Sarah did in May 2007 ). Unfortunately, like many contestants on the show nowadays, Rachel was too terrified of the 10% disaster chance and decided to call it a day.

Haleem (4 November 2005):



Deal-Eye still not quite getting a win, though at least it draws this time. The increase from £7,900 at 11-box to £18,500 at 8-box after taking out the £100,000 (and two smaller values) is really strange.

Apparently there was no final offer for this game.

Tally after 5 games
0 wins
3 losses
2 draws
Deal-Eye: £82,900
Players: £100,400