This is the forum.

dear players,

it seems i have some time on my hands to fix some things here and there.

let's start with bugs.

- games getting stuck. i think i know how to fix this
- comcon character join bug. i can't remember what it was but i think it's there
- VP prices not evolving. or are they evolving now?

any other know bugs?

thanks

neko

I do not believe the ship stats are evolving at all! They stay the same no matter what happens. If the combat reports are to be believed, that part SAYS they are updating the stats for ships and races, but, they never do as ship values are a constant.

Prices do seem to change for those games with custom builds.

Any chance of "fixing" the multiple ships per player problem? Perhaps the code is not in place to do so, but, signup seems to be set up to allow multiple ships per player.

Not a bug per se, but, when joining a game, sometimes, the creator says things like only use a fireball. That's great, but, I have accidentally joined many times with Lone Wolf, which is the default in the selection box. I wish this was changed to "Select One" so that you can not accidentally pick lone wolf.

neko wrote:

Hey Neko, glad to see you working on the site again!

Regarding the issue above, I think the prices of different components for custom fighters are evolving as they are intended to, more or less (there are some distortions, but they're due to some items being much more popular/effective than others). What are not changing are the ratings of the fighters (and the factions for custom fighters). Since the ratings effect scoring, this can distort player's incentives for picking various ships (Comcon fighters, for example are especially popular for custom designs because of that faction's low rating).

The other main bug I've seen is that the asteroid field icon (showing how dense a field is set) does not display on the "Join a new game" page. It used to not show up on the "Running games" page either, but it looks like you've already fixed it there.

A trivial bug is in the AoW turn report, in the "Invasion Results" section. You get text like: "Bounce: Russia (14) bounces on Germany with 12 units, at least 14 were needed to try." The number in parentheses is misleading because that's the number of units that have already claimed Germany, not the number from Russia. It can probably be left out entirely, since it's repeated later on.

Also regarding bounces, there's something that I don't know about the implementation which might contain a bug: If more than three territories attack and defeat a common neighbor's defenders and have the same number of survivors, do they all have the same odds of claiming the territory? That would not be the case if each "Random Bounce" check is always a 50/50 random coin flip. The territories that get checked first would end up having poorer odds than the one that is checked last. Since the order of the bounce checks is arbitrary I doubt this has ever been exploited by anyone, but it might be something to look at. The solution is to keep a count of the random bounces/claims in each territory, and the first one will be 1/2 chance, the next 1/3, the next 1/4. This works out to correctly give each attacker an equal chance.

The only other issues I have are more game-design suggestions, rather than bugs:

The VP collision detection algorithm is often not what people expect it to be, and it might scare away some newbies who never learn it. We might get away with keeping the damage algorithm the same, but I think any move that passes through a hex with an asteroid should count as a collision of some sort (even if it's just a zero damage one). The new collision probably helps a bit, but it's not as nice as having a more intuitive system.

A slightly related issue is how missiles behave if a player ends a turn on the missile's starting position, which can be rather unexpected. I think we discussed some ideas about this a while ago in the forum, though I don't remember the outcome.

And of course it would be fun to add new missile or grenade types. An inertial grenade that didn't always explode on the turn of it's launch, but only when it had a target (missile or fighter) in it's explosion range would be a lot of fun (and a lot more useful than the current one). A "mine" that would slowly seek towards a target would be fun too (give it one or two thrust in any direction, using something like the smart-missile algorithm I posted about a while ago).

Anyway, I hope these reports and suggestions help!

neko wrote:



- Games can't be started with multiple fighters per player.

- Maybe a delete game button. Only available for the player who setup the game and the game didn't start yet. Just to remove games that are made with the wrong settings.

Probably not very important, but I get a

Warning: Invalid argument supplied for foreach() in /web/htdocs/www.takeaplay.net/home/forum.php on line 231

on my previous post. I imagine it is related to the fact it does not have any reply yet.

sfatula wrote:


I see! Thank you!

sfatula wrote:


I am afraid this thing will not be in this specific round of fixes. The code was originally meant to allow for multiple ships per player, but from some point on I decided against this, and coded the "one fighter per player" way. It may or may not be easy to do this, and at the moment I haven't the faintest idea...

Thanks a lot for the reports!!

BlckKnght wrote:


Oh. I didn't know this.



Or this...



I can't remember if at one point I fixed this, but I believe that, as you say, territories with a lower ID still have a marginally higher probability of getting the territory. I'll have a look at it and see how easy it is to fix this. It should need an extra while-if-then-else-wtf loop, and some extra dummy variables




I agree with you. I think it's better to have an "all or nothing" collision rule. Either full damage, or nothing.



Thanks for the reminder, I'll look for that discussion.



Not this time, sorry

ivlivs wrote:


Probably. Thanks!

How 'bout no emails for games in which I have been eliminated from! Some go on for weeks.

If this is a bug or not i still would like to mention it.

Please change the scoring methode.
If a player makes a hit then give him/her points related to the amount of damaged that is done instead of points to whom makes the kill in the end. Often others players shoots the ship to wreckage but but a few bits of armor left. The next player gets all the credits for the heavy work.

One more minor bug (which I think I've reported before, but perhaps not somewhere obvious):

On the "Ended games" page, all AoW games that end while you have switched to automoving are listed above the most recently ended games. For players like me who tend to use the "Surrender and enter Automove" diplomacy option in games that are hopeless, that means we have a lot of clutter on the top of that page. I think it's good to have such a warning on the main "Running games" page, but it's useless for ended games.

sfatula wrote:


Didn't know this. Duly noted.

AshayRey wrote:


You have a point there. I will be re-thinking the scoring method, separating the game points (which determine who wins a game) from the ELO points (the overall stats).

BlckKnght wrote:


The whole presentation of active and ended games will be re-organised. I hope the info will be clearer then.

neko wrote:

Nifty!

I'm not sure if this is part of your plan already, but I'd really love to be able to browse all of my previous games, not just the last 50 or 100.

One reason I've wanted this is to go through my 1v1 AoW games to see how strong the statistical evidence is for some territories being enormously better than others. Because of all the VP I play, my last 100 games don't give enough data to make any sort of conclusion (some starting territories will probably not appear at all in the data set).

That brings me to another suggestion (not a bug, but an improvement opportunity):

Consider changing how AoW places the players initially to try to maximize their distance from one another. Especially in games with defending Neutrals, starting two territories away from an opponent is often no fun (the game usually hinges on a single battle that occurs in turn 5, and if there are other players with lots of room to expand in, they'll eat you up later).

Another option is to do some statistics like I wanted to do myself and then try to pick "fair" starting positions for each game. A way to do this might be: For each game with a given setup (e.g. 2-player, no defending neutrals), average the points scored by players starting in each territory. So Scandinavia (one of the better territories) might average +4 points per game while Mexico (one of the worst territories) might average -5. What the game would try to do is pick sets of starting territories with relatively small differences in value.

Or one final alternative: You could leave the positioning random, but scale the points awarded at the end of the game by the value of the starting territories (this could be very similar to how VP fighters are weighted by their value).

I have another improvement suggestion that's not really a bug:

I'd like the VP game setup interface to let me select whether pre-made or custom fighters are allowed in a game, separate from the rule level (simple/standard/advanced/experimental) I pick. This would allow, for example, a game to be played with standard rules using custom fighters, or alternatively, a game with advanced rules to require a the use of a standard design.

I'd especially like the ability to block premade fighters (or to put it differently, require custom fighters) in games without altering the budget (and thus making the game not-for-stats).

I realize that this could potentially cause problems with the rating system, as playing without the standard designs (or with only them at high rule levels) could unbalance the the custom ratings as compared to the non-custom ones. I think this is already the case though, since the lower rule levels already restrict fighter choices. While the ideal solution would probably be to have separate fighter ratings for each situation (like the scenarios have) I think we could easily do without them and let the ratings go where they will. (I don't think there's likely to be any systematic difference in performance between a faction's custom fighters that are designed to go against other custom fighters and the ones that are designed to fight against standard designs too.)

A slightly trickier technical aspect of this issue is that the custom fighter design UI might need to be updated if custom fighters are allowed in games using basic rules (where there can be no missiles, rockets or IGs and no dual thruster systems). I think the simplest way to go for this would be to block custom fighters from basic rules games only, and let them be optional in all of the higher rule levels. The only remaining issue would be validating that the game setup was legal when it is created (e.g. you can't create a game that bans both custom and pre-made designs).

- forum newline bug
- forum edit post feature

BlckKnght wrote:


I would like to re-think carefully what is meant by "simple", "standard" and "advanced" rules first, to provide game levels that really make sense to play separately. I am thinking of "simple" games having no asteroids, like on an empty map. Removing the collision problem would allow a simplified interface, showing only one position per fighter. This would make the game much more user-friendly, I believe.

I am opening another thread on a related issue (rules overhaul).

Then, I think I might like to invest some time on the scenario creation editor. There is one for superusers already, but maybe having one for players to enter their own scenarios, and superusers to just "confirm" a scenario as published, would be better.

This is not exactly what you were asking for, but it goes in that direction, I guess (more diversity). However, I will be thinking of some way of having some degree of fighter choice in scenarios.

neko wrote:

It would be nice if the forum actually noticed what post I'd read, rather than requiring me to always hit the "Mark all as read" button.

neko wrote:

A scenario editor sounds exciting! Regardless of how it applies to my suggestion, I hope you'll make it available.

What I was asking for could perhaps be fit in with the scenarios, as long as there's an "Any Custom Fighter" option for the fighter selection. That's what I really want, to be able to play a game (with Advanced rules, if necessary) that only allows custom fighters (that is still for stats). The rest of my earlier post was just extending that idea to its logical conclusions.

The smiley parser in the chat box can only handle one smiley of a given type on each line. Additional ones either show up as a different smiley (e.g. the regular smiley displays as the second time) or it shows up as plain text.

(Does the forum have the same issue? Lets test: From the preview, it seems not.)

BlckKnght wrote:


Thank you, right. I should make that function recursive.

This is even less of a bug than some of my other suggestions, but I can't resist:

I hate , hate , *HATE* the AoW diplomatic model.

Very often in the middle part of a 4+ player no-intrigue no-neutrals game you'll end up with everyone set to "2-way draw" and there's absolute chaos because nobody can tell who they're allied with. Often it just ends up with the players who had short but heavily defended borders with each other backstabbing their nominal "partners" (who's borders were left lightly defended) and drawing with each other.

To make alliance offers more explicit, I think each player should have a separate diplomatic status that they can set for each other player. The results would be shown in the same grid as the historical number of attackers. So if I'm red and I want an alliance with the blue player against our mutual enemy green, I could set my diplomatic setting towards blue to be "draw" while leaving my status with green at "no mercy" .

Similarly, you can surrender to a certain player while still expressing that you will be fighting against somebody else, and or you can accept some surrenders but not others. Switching on automove would be available, and it would offer surrender to everyone.

This wouldn't eliminate the potential for backstabbing entirely (which I think is good, at least in games with non-binding diplomacy), but it would make the action explicit, as defection from one alliance to another would require the changing of diplomatic settings.

- VP movement reconstruction algorithm "bug"

In an AoW game, when the last player to move in a turn is automoving, the game will not check to see if it should end for diplomatic reasons (I'm not sure if it will skip the check for ending due to 31 territories too). This can often happen in two player games (or when all but two players have been eliminated or been on automove for a while) if one player surrenders (and enables automove) and the other player does not accept the surrender until after they submit their next turn. The second player can often get their move submitted during the time the automove system delays before entering its move.

I just had this happen in game 6532 (http://www.takeaplay.net/sup_browse_game.php?game_id=6532 ). Pnzrkitty surrendered on turn 7, but I didn't notice it when I submitted my orders for that turn. When I saw the results I realized that he had not attacked at all, and noticed his new diplomatic status. I accepted his surrender on turn 8, but my manual were submitted before his automove orders. Rather than the game ending on turn 8 as it should have, a useless turn got processed before the game finally ended at the start of turn 9.

I realize that there are good reasons for the delay in executing automove orders (otherwise a game with all automoving players could eat the servers CPU). This seems to be a wrinkle that got missed in the proper handling of it though.

Hi, I found the link "create or join a new game" on top of http://www.takeaplay.net/sup_rules.php to be broken.