MrBenn wrote:The neutral on Durham makes sense, and I'll have to decide whether to start it with 2 or 3 neutrals in due course.
if there is only one preset neutral on the board, then durham is where it will do most good.
DJ Teflon wrote:My issue with the +1 for 4 scenario is that the player going first is pretty certain (I'll work-out and edit in the percentage later) of getting +1 in the midlands and +1 in the south (quite possibly +2 in both). Most players will immediately use this advantage to make sure their opponents don't get as many, or any, +1s on their first go. Big advantage to the first player (or team). (Rolling 7 or 8 against 3 is much more likely to win that 6 v 3). Player, or team 2 may lose the will to live.
MrBenn wrote:The Midlands and South build-your-own bonuses could be changed to +1 per 5, which reduces the chance of getting them on the drop; I'm not keen on making them much bigger as I want there to be a chance of getting them at a point that matters to the game.
i agree with dj's analysis. a likely 50%+ chance of player 1 dropping a bonus makes player 2's position unnecessarily difficult. a solution that retains the spirit of build-ur-own bonuses is to change them (and the wording on the road sign) to +3 for every 5 counties except the first 5. under this bonus system, at least one player is extremely likely to start with either 7 or 8 midlands counties and be in a position to hold 10 counties after a few turns if desired (defending them is another matter!).
MrBenn wrote:The coded starts will only affect games with up to four players, although I'm contemplating reducing this to three, and not coding the North West.
good move! 4 sets of start positions will be totally ineffective at ensuring that yorkshire is divided at the start of 1v1 games, since player 1 might have the set that includes north riding, while player 2 might have the only set that has no yorkshire ridings, meaning that the other 2 ridings are allocated randomly. 3 sets of start positions that consist of one yorkshire riding and one thames valley county only will be better for this purpose.
MrBenn wrote:The Midlands and South build-your-own bonuses could be changed to +1 per 5, which reduces the chance of getting them on the drop; I'm not keen on making them much bigger as I want there to be a chance of getting them at a point that matters to the game.
i agree with dj's analysis. a likely 50%+ chance of player 1 dropping a bonus makes player 2's position unnecessarily difficult. a solution that retains the spirit of build-ur-own bonuses is to change them (and the wording on the road sign) to +3 for every 5 counties except the first 5. under this bonus system, at least one player is extremely likely to start with either 7 or 8 midlands counties and be in a position to hold 10 counties after a few turns if desired (defending them is another matter!).
The problem with this is that it gives too much emphasis to 1v1 games, when I'd prefer the map to be better for larger-player games... In an ideal world we could give different bonuses to byob's in 1v1s than 8p games, but some form of compromise is going to be in order here... There is some merit in something like +1 for every 2 above 5 for example, so 5 terrs=+1, 7 terrs=+2 etc.. I like that part of the suggestion
iancanton wrote:
MrBenn wrote:The coded starts will only affect games with up to four players, although I'm contemplating reducing this to three, and not coding the North West.
good move! 4 sets of start positions will be totally ineffective at ensuring that yorkshire is divided at the start of 1v1 games, since player 1 might have the set that includes north riding, while player 2 might have the only set that has no yorkshire ridings, meaning that the other 2 ridings are allocated randomly. 3 sets of start positions that consist of one yorkshire riding and one thames valley county only will be better for this purpose.
With 4 positions, then each player will be given 2 of them in a 1v1. In a 3p game, each player will be given one, and the other terrs in the start position will be allocated randomly. The coloured 88s on the last update showed my suggested groupings - I put a fair bit of thought into ensuring that any combination will have a broadly even start in the North... if p1 gets 2 Yorks terrs, then p2 has more terrs bordering Yorks to try and break it, should p1 manage to take all three Ridings. The same access theory was applied to the North East...
Despite my personal views about the role of the Beta phase, I'm beginning to think that the exact byob values should be tweaked and refined while the map is in Beta play (which feels like it might never happen at this rate).
PB: 2661 | He's blue...If he were green he would die | No mod would be stupid enough to do that
MrBenn wrote:The coded starts will only affect games with up to four players, although I'm contemplating reducing this to three, and not coding the North West.
good move! 4 sets of start positions will be totally ineffective at ensuring that yorkshire is divided at the start of 1v1 games, since player 1 might have the set that includes north riding, while player 2 might have the only set that has no yorkshire ridings, meaning that the other 2 ridings are allocated randomly. 3 sets of start positions that consist of one yorkshire riding and one thames valley county only will be better for this purpose.
With 4 positions, then each player will be given 2 of them in a 1v1.
are u sure that this is the way excess positions are allocated? i'm not saying that i know better, just that i seem to recall evidence from new world which contradicts the 2-2-0 allocation, but i'm unsure myself and don't have time to look it up before i go away for the weekend.
The hold-up we have currently is with thrashing-out an acceptable system for the 'build your own bonus' scheme.
Bonus Drops with the Current Version Thanks to MrBenn's fantastic calculator (assuming it gives us a close answer to the true probability of dropping specific bonuses), we can establish the following:
MrBenn's Proposal to not have codings for 4-player games We would have 100% chance of dropping a +1 for the first player (on 'build your own') in both the south and midlands. This would be a massive advantage to the first player. I really think I should insist on coded starts for 4-player games, a +2 advantage to the first player is a no-go, sorry.
Bonus Drops with MrBenn's Proposal to remove coding for the North-West MrBenn's calculator establishes the following:
Conclusion of Analysis of Probabilities for Current Version and MrBenn's Proposed Changes
Current Version Would only really have acceptable probabilities of dropping a 'build your own' bonus if we had a +X for 7 territories system. Even a +X for 6 territories system is entirely debatable.
Proposal to not have coding for 4-player games Would give 100% probabilities of players dropping a bonus, which is unacceptable given the big advantage to the first player/team.
Proposal to not have coding for the North-West This would affect the probabilties to a small extent and would mean that even +x for 7 territories is very questionnable.
Possible Alternative System Subjectively, I really would like to see the 'build your own' bonus stay for this map and I am keen to help thrash this out for it to be workable.
Objectively, it needs to be amended if it is to work.
Here is a possible solution given that MrBenn would prefer the 'build your own' system to involve bonuses for the minimum no. of territories possible:
Coded Starts in all 3 'regions' for up to 4 player games:
In the North, Yorkshire is the area where coded starts are most vital (given the neutral in the NE). Therefore we have coded starts for Yorkshire and Northumberland. The probability of dropping the NW bonus is ok - only 0.13% for the first player (in a 4-player game - less than either benchmark mentionned above).
In the Midlands, coded starts in the West Midlands (where there are 5 territories) eliminate the very slim chance of bonus-dropping there and eliminates the possibility of players getting a cluster of territories in the NW and West Mids (which isn;t really a big danger). Basically,if we have coded starts somewhere in the Midlands, we can get acceptable probabilities of dropping a 'build your own' bonus'.
In the South, coded starts are most suitable on the Thames region (to eliminate the small probability of dropping this +4 bonus).
If this were implemented then we would have the following probabilities of dropping a 'build your own' bonus:
Summary of Suggestion I would propose that there are 12 coded starts - 4 in each in each of the North, Midlands & the South and that the build your own system is +X for any 7 territories (e.g. +3 for any 7). +X for 6 may have a remote, outside chance of being arguable, possibly.
Comparisons with standard bonuses are dangerous, because you're not comparing like for like. It would be better to make comparison with a map like Berlin that already employs build-your-own bonuses.
PB: 2661 | He's blue...If he were green he would die | No mod would be stupid enough to do that
Very true MrBenn, a well made point, especially as we are talking a +1 bonus in this case and, more importantly, 'build your own' bonuses would be more difficult to hold, generally.
Berlin's most likely dropped bonus is the French zone - 8.93% chance for player 1 to drop.
If we take that as a baseline, then, if we accept the figures I gave above, we would be looking at a +X for 6 system with my proposal and +X for 6 would remain debatable for the current system or MrBenn's proposals.
P.S. Apologies for slowness in providing the above analysis
DJ Teflon wrote:Very true MrBenn, a well made point, especially as we are talking a +1 bonus in this case and, more importantly, 'build your own' bonuses would be more difficult to hold, generally.
Berlin's most likely dropped bonus is the French zone - 8.93% chance for player 1 to drop.
If we take that as a baseline, then, if we accept the figures I gave above, we would be looking at a +X for 6 system with my proposal and +X for 6 would remain debatable for the current system or MrBenn's proposals.
P.S. Apologies for slowness in providing the above analysis
i have also noticed the when the french zone is dropped there is a big chance that the check points bonus would be held by the other player or at times one of the larger zones. Thus it evens out
How about this for an update... I'm actually fairly happy with this, although have been debating whether the BYOB should start at 6 or 7 terrs (settled with 7 to appease the 1v1'ers). To mitigate this I've upped it to a +2, with +2 for each subsequent 3 (so 7 terrs=+2, 10 terrs=+4, 13 terrs=+6, 16 terrs=+8) I was umming and ahhing about whether to make it +1 per 2, but went with +2 for 3...
PB: 2661 | He's blue...If he were green he would die | No mod would be stupid enough to do that
I was just thinking, it's a religious symbol aswell and I figured CC maps should be made so that no one can 'take offense' or feel less welcome or anything Then again, if it actually suits the map, as you have explained... England is also Christian for like, what... 95%?
(I just looked it up: 71,6% of British population was Christian in 2001. However, if you exclude the non-religious, it's a little over 93%.)
Winner of "As Easy As 1, 2, 3! - Africa I", "Championship Series: British Isles", "1v1 Battle to Rule Doodle Earth 2", "Connect 4 (Restarted)" and "Blind Fold Buddy - BeNeLux"
Setting the 'build your own' to start with 'for 7' is safe, drop-wise in 4 player games. And, the +2 system makes sure the feature is a major aspect of players' strategies.
I think we can safely say that this system would be entirely workable for 4-player games (including 2 v 2).
In the 2 or 3-player game we would still have significant probabilities of dropping a +2 for 7 in the Midlands - 35.0%. The probability is 11.83% in the South - so possibly adjustments needed for 2-and-3-player games. If we go down the route of adding more coded starts this may not be too satisfactory (my initial calculations suggest we would need lots and lots which may not be preferable).
If I were to code the whole map into 3-player starts, then In 3p games, no-one would get any bonus to start and Middlesex would start neutral (through virtue of being unhandoutable rather than coded neutral). I could code Durham as an underlying neutral for 4p+ games, and 2p games if the blue set doesn't get dished out.
In 4p games, the odds for the Midlands are 5.28% for +2, trace% for +4; In the South the odds are 2.33% for +2, trace% for +4. There is 1.05% chance of getting Yorkshire, and less than 1% for any of the other discrete bonus areas. For >4 player games, the odds are significantly lower, and in 7/8 player games a physical impossibility.
This leaves the 'problem' of 1v1 games, which are well-documented to be drop/dice/1st-turn dependent in any case.
1v1 Games
With the three starting positions in a 1v1, each player would be given one of the red/green/blue sets. This leaves 14 or 15 territories to hand out (depending on whether the blue ones are served up with the underlying neutral on Durham).
Each player will have 6 Midlands terrs, with 6 left in the pot. There is a 95% chance of any player getting one additional Midlands territory for a +2 bonus, but only 4.7% that they'll get 4 of them to get up to a +4 bonus. An even distribution would see each player getting 8 Midlands terrs - so to break the opponents +2 bonus, you'd have to kill 6 with a deployment of 5.... which I think is going to be the best balance without screwing up other game-types.
It's a similar picture in the South, with each player having 5, with 6 left in the pot. here there is a 70.6% chance of getting the +2 for any 7 bonus, and 0.2% chance of getting the +4 for any 10. If one player has a "good" Midlands drop, then the other player is more likely to have a "good" Southern drop (out of the 11 territories available in those areas) , and I'm satisfied that this is as balanced as we could possibly get, without changing the system to something completely different.
PB: 2661 | He's blue...If he were green he would die | No mod would be stupid enough to do that
I think two of the three coded starts (which are now the three colours of 88s) would just go to either player, and the third group would be divided between them and neutrals (as all territs in normal maps would). At least, that's what I make of it Edit: fastposted
Winner of "As Easy As 1, 2, 3! - Africa I", "Championship Series: British Isles", "1v1 Battle to Rule Doodle Earth 2", "Connect 4 (Restarted)" and "Blind Fold Buddy - BeNeLux"
plain quick easy gameplay superb graphics what do you want more stamp and go live
if you really want something to do maybe the frame around the legend a bit too mrbennish try something else than a white rounded one this is a very open moving gameplay bonus are hard to held any impassable woods? but really this one is ready for playing
De gueules à la tour d'argent ouverte, crénelée de trois pièces, sommée d'un donjon ajouré, crénelé de deux pièces Gules an open tower silver, crenellated three parts, topped by a apertured turret, crenellated two parts
I didn't knew it was like English road signs as it have a precise graphic reference then it should remain as it is maybe a pole or a small rotation as road signs always are... or it is just fine like this
De gueules à la tour d'argent ouverte, crénelée de trois pièces, sommée d'un donjon ajouré, crénelé de deux pièces Gules an open tower silver, crenellated three parts, topped by a apertured turret, crenellated two parts
Games with 4-Players+ (incl. 2 v 2) As stated before, this is now fine. It looks like, the current suggestion is not to have any coded starts for 4+ players (just the neutral on Durham). Again, this is fine. Probabilities of landing either a 'build your own' or a 'normal' region bonus are sufficiently small.
3 Player Games The coded starts remove any concerns. This is fine.
1 v 1 Games It is difficult to work-out exactly how often player two would drop a bonus in the south to counteract when player 1 drops in the Midlands. I am still concerned about this.
Would it be possible to have the blue territories (in the 3-player) coded as neutral in 1 v 1? If enough of the blue Midlands and South territories were coded as neutral for 1 v 1 this would eliminate bonus drops (they could be coded as 2 neutrals instead of 3 to encourage players to 'build their own? If this is possible and acceptable, I would suggest all blues in the Midlands coded as neutrals and three of those in the South.
Alternatively, and along similar lines - if there were no coded starts but one neutral in each of the 9 regions then the chances of player 1 dropping a +2 for 7 bonus would be 11.22% in the Midlands. If an extra neutral were in the Midlands, as opposed to the South (i.e. 4 overall - 2 in one of the regions), then this probability would drop to 7.15%.
Last edited by Teflon Kris on Fri May 29, 2009 8:41 am, edited 1 time in total.
It looks like the only concern at the moment is 1v1 games (which are inherently biased towards whoever goes first on any map/settings)
DJ Teflon wrote:1 v 1 Games It is difficult to work-out exactly how often player two would drop a bonus in the south to counteract when player 1 drops in the Midlands. I am still concerned about this.
Would it be possible to have the blue territories (in the 3-player) coded as neutral in 1 v 1? If enough of the blue Midlands and South territories were coded as neutral for 1 v 1 this would eliminate bonus drops (they could be coded as 2 neutrals instead of 3 to encourage players to 'build their own? If this is possible and acceptable, I would suggest all blues in the Midlands coded as neutrals and three of those in the South.
With the current set-up with three starting positions, each player would get 4 or 5 territories in addition to their starting group (the leftover territories from the unassigned group).
There will be 6 territories in each of the Midlands and South to get handed out, so there is always some chance of both players getting an extra territory in each area to get up to the 7 terrs bonus. In the worst-case scenario of p1 getting 5 additional Midlands territories, then there is a 50% chance that p2 will have the other one.and 100% chance of them having at least 1 additional territory in the South, and a 26% chance they'll have 4 or more Southern territories.
I'm satisfied that I've done all I can do to make 1v1s as balanced as possible, without adversely affecting other gametypes. If it transpires that there is an overwhelming disadvantage once the map is in play, I think we'll have to revisit it then.
PB: 2661 | He's blue...If he were green he would die | No mod would be stupid enough to do that
I appreciate that it wouldn't mess-up on-going games if start positions for 1 v 1 were amended in Beta. Also, the chances of dropping +4 for 10 in the Midlands is an acceptable 1.5%. However, I would predict that player one would often have a huge advantage from the following scenario:
Dropping 2 x +2 bonuses in the Midlands and the South in 1 v 1 With 14 territories to be handed-out, the chances of player one getting 1 of the 6 available in the Midlands and 2 of the 5 in the South are high (although difficult to calculate).
In this scenario, player one would most likely be able to deploy his 7 troops strategically and prevent player two getting either the two or one bonuses (s)he may have dropped. Certainly, player one is most likely to retain a bonus advantage going into, and beyond, round 2.
I like your ideas as far as the bonuses go for this map, it just looks like it would be hard to maintain any territory. There appears to be alot of outlying borders.
