 |
| Book'em Dano 2003-03-07 17:50:11 | Opera 6's winning percentage is now at 45.576% while Mozilla's is 45.557%! How can this be? Will it last long, or will the bots return order to the boards? Stay tuned only time will tell . . . . |
| ß 2003-03-07 18:14:27 | also, o7 is experiencing a great boost in its per centage and it on its way to 41%. watch out dillo... <font size=1>now i think of how ironic that actually is...</font> |
| Stephen Brooks 2003-03-08 10:29:58 | Those two (O6 and Moz)'s percentages have been eerily close for over a month now. I was even wondering if there was a thing going whereby a load of Moz users were asked to play whenever O6 got ahead. Today's weather in %region is Sunny/(null), max. temperature -99999°C |
| ß 2003-03-12 11:07:26 | now there is only a one thousandth of a per cent difference between the two, with o6 still in the lead. seems to be that moz won a bunch against o6. you might just be right stephen... |
| ß 2003-03-12 12:56:04 | moz now leads by a hair at 45.506%, to o6 at 45.505%. |
| Book'em Dano 2003-03-13 17:31:50 | From what I'm seeing right now O6 is playing a minimum of 3 pieces on several boards but nothing beyond that. Hmmmm . . . . |
| Stephen Brooks 2003-03-14 08:25:32 | The only way O6 can defend against those tactics is to have a genuine player on watch who can complete those games. OTOH you could do the same to Moz. BTW Links has a pH34R50m3 win percentage, and has maintained it around 48-49% as its games-played have increased. Today's weather in %region is Sunny/(null), max. temperature -99999°C Hmm do you think Mozilla's supremacy is gradually being challenged? Out of the 10 most common game-combinations on the bottom right of the main screen, only 2 are now being won on the whole by Moz. They are the most common two, but even so, the Moz-O6 score is only 32 games away from being equal. No sign of MSIE6-Moz games moving though. [This message was edited by Stephen Brooks on 2003-Mar-15 at 15:26.] |
| kpu 2003-03-15 11:43:28 | "BTW Links has a pH34R50m3 win percentage, and has maintained it around 48-49% as its games-played have increased." Why, thank you Stephen  . I generally play with mozilla until the board gets crowded when speed matters. Then I switch to links in graphical mode and usually get a piece in every other move. |
| Stephen Brooks 2003-03-15 14:26:45 | Well that probably explains it. The Links games have sort-of been filtered down to the ones when you're deliberately playing fast. Today's weather in %region is Sunny/(null), max. temperature -99999°C |
| Book'em Dano 2003-03-15 16:44:12 | Moz is behind O6 again . . . by .002 points. Maybe we'll have another chance of seeing if Stephen's theory proves correct. |
| anon 2003-03-15 20:49:28 | <SARCASM>Is Stephen's theory basically that when Mozilla is below Opera, Mozilla will eventually be above Opera later? You mean that the browsers actually switch positions over time? Wow! Of course, no explaination like statistical randomness can explain these small changes. Could the same logic be applied to say that Opera gets more players when it is behind?</SARCASM> |
| Stephen Brooks 2003-03-17 12:46:16 | Heh. I condensed the explanation of my argument to the point that others could only 'get' what I meant if they assumed I wasn't saying something trivial. Here's the extended version... Firstly, I noticed that Moz and Opera 6's win percentages were _unusually_ close, which means that I got the impression that they were closer than if they were produced by two independent statistical processes. I was seeing the difference between them being no more than 0.050% for a very large proportion of the time. Typically statistical fluctuations scale by 1/sqrt(N) where N is the number of trials, so here N=40000ish so 1/200 = 0.5% is the sort of variation you'd expect, even between equally-skilled teams. There's obviously always the possibility that it's just a plain coincidence, but I thought it rather unusual. So far the situation is symmetrical and all I've implied is that there is some interaction between the two teams, for instance more people from the losing team playing (or perhaps the good reserve players coming out of their cupboards to save their favourite browser - whatever). The reason I thought that it was _Moz_ 'monitoring' the scores was that it also seemed that, say 85% of the time, I saw Moz's percentage as just a _little_ above Opera's. The rest of the time Opera was ahead by an amount that varied much more, so I reckoned that some Moz were playing just for the purpose of getting the % a bit in front of O6, whereas the O6 hoarde just pillaged the scene at intervals regardless of the percentage. The only way I could imagine such organisation would happen is if there's a BBS somewhere with a small Moz clan who want to regulate the scores. OK, perhaps it's a bit of a paranoid theory, but I thought it was fun  Today's weather in %region is Sunny/(null), max. temperature -99999°C |
| Book'em Dano 2003-03-17 19:46:52 | Stephen, it would seem to me that the number of players on the boards are decreasing in comparison to previous months (i.e., look at the increasing times between games played and games finished). Consequently, another alternative to "The only way I could imagine such organisation would happen is if there's a BBS somewhere with a small Moz clan who want to regulate the scores." is that a lot of the Moz clan are very familiar with bot usage. How else does one explain the 4,000+ game spread between the two browsers. [This message was edited by Book'em Dano on 2003-Mar-18 at 5:41.] |
| kpu 2003-03-18 15:23:28 | Um. . . maybe more people use Mozilla? How does one explain the 5000+ gap between O7 and O6? O6 players must be more familiar with bots. Yes, there is a general decrease in playing. However, a decrease in playing does not mean much to the stats. It just takes more time for things to happen. I guess you're trying to say that the time interval during which mozilla was behind is small. Compared to previous incidents, this is MUCH longer. Look at the threads where somebody brags about their browser beating mozilla where the next post has somebody refuting the claim. You are also making a rather large conclusion based on one data point. |
| Book'em Dano 2003-03-18 18:00:32 | One possible reason for the difference in the number of wins between O6 and O7 is that O7 was not officially released until January 28, 2003; hence it hasn't been played as long. I'm not sure when the boards made the distinction between the two (i.e., O6 & O7), maybe it was during O7's beta stage in November/December, I'll leave that to Stephen, but I do remember a lot of O players complaining about the split. Along those same lines Mozillla's player count hasn't been split because there have been no new major releases in the product to warrant a renumbering (i.e., 1.3 versus 2.x), but then again all of Phoenix's wins count towards Mozilla's. I honestly don't think there are a lot of Phoenix players but you never know. I'm not complaining mind you (remember its a war), just stating what I recall since I've been playing. And yes some O players do know about bots, but I for one think the Mozilla players were more flagrant about it in times past. In fact, I remember one who got mad when I recognized the pattern of play and locked the board and then they complained they couldn't finish developing it. But I digress, anyone want to play? |
| Stephen Brooks 2003-03-19 08:59:56 | Who's KPU replying to two posts above? I wasn't making a conclusion on one data point, I said I'd noticed a trend over about the past month for leads over Moz in win%age by O6 not to last as long as leads the other way around. I guess it could also be a bot, or coincidence too. Today's weather in %region is Sunny/(null), max. temperature -99999°C |
| Stephen Brooks 2003-03-19 12:58:15 | LOL not that it's relevant now anyway... Moz has pulled waaay ahead of O6 Today's weather in %region is Sunny/(null), max. temperature -99999°C |
| kpu 2003-03-19 16:33:33 | I was replying to Book'em Dano in the post directly before mine. |
| Stephen Brooks 2003-03-21 15:03:51 | OK, no problem. I found someone complaining about Lynx (or Links) not being able to view the 'dsiplay last win' screen, so I (1) put alt-text for the mini icons there, which I had left out due to some oversight, and (2) made a non-popup version so maybe text-mode browsers without JScript will find it easier. Today's weather in %region is Sunny/(null), max. temperature -99999°C |
| ß 2003-03-23 10:26:49 | the last wins thing works well enough in lynx, but the empty rows make it look ugly. they need to have an empty-row alt-text, but otherwise it's perfect. n.b. the alt-text thingy would be nice on the main page, say, by using a single letter/symbol for each browser and adding a legend at the top of the page if the php detects a text browser. the only problem i see with this is with someone using graphical links, but that just sounds cool to me. peace now  |
| Stephen Brooks 2003-03-23 15:36:37 | I think that might have been my reason for not putting alt text in there to start with. I avoided alt text altogether on the index page because with Errr 12 boards each of something like 10x8 cells, that's 960 copies of " alt=''" (7 chars) as overhead before you've even written the content of the alt text. That amounts to 6-7KB per page load, and I was trying to trim the front page as much as reasonable. But I'll try to fix the blank-rows alt text... Maybe put "----- ----- ----- ----- ----- ----- ----- ----- ----- -----" for a whole row. Today's weather in %region is Sunny/(null), max. temperature -99999°C |
