Recommend
9 
 Thumb up
 Hide
30 Posts
1 , 2  Next »   | 

Bug» Forums » General

Subject: Merging groups allowed in placement but disallowed in growth rss

Your Tags: Add tags
Popular Tags: [View All]
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
Nick, after playing (and enjoying!) 19 times, I have a couple thoughts/questions about merging groups:

1. In our first few games, we messed up the rule about growing after eating when a merger would happen. In reality, the eating does not occur if a merger is inevitable. In our "unintentional variant", the merger occurs if the resulting group is still not bigger than the biggest group which existed before eating, i.e. just like how merger during placement is OK if the resulting group is not bigger than the current biggest group.

This confusion of ours made me wonder about the seeming inconsistency in the placement and growing rules: did you playtest allowing merging during eating if the resulting group wasn't too big? And if so, evidently you preferred the existing rule (that growing must be by exactly 1 stone); can you say why?

It seemed like our (mis)interpretation worked OK too, and would then seem more consistent with the placement rule (albeit perhaps less simple?).

===

2. Thematically, growing by only 1 (with no merging allowed) after eating makes some sense, if each group represents a "bug". After all, in "real life" individual bugs don't usually merge together into larger bugs! "Real life" bugs are individuals which grow as separate entities, rather than merging into each other. Which made me wonder whether I was confused about the placement rule, and perhaps merging was also disallowed there. Merging your bugs seems an unfortunate theoretical "thematic disconnect", and having mergers allowed in placement but not in growth seems a slightly annoying/confusing theoretical inconsistency.

The placement rule is in fact a bit coy about this, not mentioning mergers explicitly at all! From one sample game in the animated GIF of 6 sample games, I saw one merger happen, so evidently they are indeed allowed. I suggest making this more explicitly clear in the placement rule! Or at least show an example of a placement merger in the rules.

And I wonder whether you playtested having placements also disallow merging? I've not tried it, but I'm guessing the game would not be as good without placement mergers allowed...?
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb

Hey Russ,

This is the rule I agonized over the most, and I'm still not sure I got it right. I did it because i suspected, but cannot prove, that gameplay is better when groups can merge during placement. However, i could very well be wrong, or at least not right enough to justify the extra confusion it creates.

My feeling right now is that, if you like it more disallowing mergers at all times, go for it. With more feedback from more players, it brings us closer to figuring out which option is really best.

N
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
We just tried a game (size 4 hexhex) with no merging allowed at all, and it actually worked fine!

When I mentioned this discussion to Anna, she remarked that in our games so far, there were in fact not many merge plays which happened in practice, and in our trial no-merge game, we only noticed one moment where one of us might have wanted to merge in reality.

I also remembered that when I was tracing through multiple games in your animated GIF of 6 sample games, I saw only 1 instance of a merge! (It could be interesting to see statistics in AI-played games how often merges actually happen.)

So in fact disallowing merges perhaps does not drastically change gameplay! And disallowing merges does seem clearly more "thematic" with the idea of individual bugs.

But presumably more tests should be done, including on size-3 boards.


PS: FWIW I was initially wondering whether disallowing mergers completely would alter gameplay because I was (mis)thinking that if little groups couldn't merge into bigger groups, then bigger groups would appear less often, because board space would quickly run out. But that reasoning seems demonstrably bogus now, because I was forgetting that of course usually the little groups get eaten and removed, and the eaters then grow.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
I agree it doesn't feel much different at beginner levels of play. My worry comes in imagining what high-level play will look like.

I believe that, as players get more skilled, the size of the largest groups that will tend to form on the board will shrink. I don't want it to shrink below four for the board sizes on which high-level play will take place (at least hexhex4), because that would strip away a lot of geometric diversity, and I see allowing merging in placement as a way to help ensure that.

(I know this begs the question about the extent I should be focusing on high-level play at all, given that most of these games never acquire high-level players, but the whole reason I design abstract games is because I love their depth, so there's a pride-of-craft thing that comes into play here).

p.s. I sent a geekmail to you yesterday about another matter. Want to make sure you got it.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
milomilo122 wrote:
I agree it doesn't feel much different at beginner levels of play. My worry comes in imagining what high-level play will look like.

I believe that, as players get more skilled, the size of the largest groups that will tend to form on the board will shrink. I don't want it to shrink below four for the board sizes on which high-level play will take place (at least hexhex4), because that would strip away a lot of geometric diversity, and I see allowing merging in placement as a way to help ensure that.

FWIW I believe we had groups of size 5 in our no-merge game.

(And we're undoubtedly among the most experienced Bug players in the world!)

I can't imagine that bugs of size 4 would not arise on a hexhex4; that would mean that bugs of size 3 never get eaten, which seems unlikely, doesn't it?

In practice, how often do you see merging happen in games? I noticed that it seems somewhat uncommon. Is it mere coincidence that in most of the example games in your 6-game animated GIF no merging happens?

Quote:
(I know this begs the question about the extent I should be focusing on high-level play at all, given that most of these games never acquire high-level players, but the whole reason I design abstract games is because I love their depth, so there's a pride-of-craft thing that comes into play here).

I certainly understand and agree with that desire!

To play devil's advocate, what if it turned out that the no-merge variant was simply deeper and more strategic, even if you see fewer very large bugs in it?

(E.g. perhaps for strong players the gameplay ends up feeling anticlimactic once someone gets a huge unkillable bug and coasts to quick easy victory, whereas when there are lots of still killable smaller bugs, it is tense and uncertain and requires careful calculation... I am not saying that this is true, but if it were, then would the desire for strategic depth or the desire to see cool diverse shaped big bugs take priority?)

Ultimately I guess only testing can decide what works best. Maybe some AI implementations can help. It would at least be interesting to see statistics on how often merges actually happen in a big archive of games. I tentatively hypothesize now (contrary to my initial hasty guess in the OP) that merging may be an unnecessary infrequently used option, and the game is just as good without merging allowed. But I dunno.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
russ wrote:
milomilo122 wrote:
I agree it doesn't feel much different at beginner levels of play. My worry comes in imagining what high-level play will look like.

I believe that, as players get more skilled, the size of the largest groups that will tend to form on the board will shrink. I don't want it to shrink below four for the board sizes on which high-level play will take place (at least hexhex4), because that would strip away a lot of geometric diversity, and I see allowing merging in placement as a way to help ensure that.

FWIW I believe we had groups of size 5 in our no-merge game.

(And we're undoubtedly among the most experienced Bug players in the world!)

I can't imagine that bugs of size 4 would not arise on a hexhex4; that would mean that bugs of size 3 never get eaten, which seems unlikely, doesn't it?


That's the thing I can't figure out! It's not an easy question to reason about. But I think there's some chance of it.

Quote:
In practice, how often do you see merging happen in games? I noticed that it seems somewhat uncommon. Is it mere coincidence that in most of the example games in your 6-game animated GIF no merging happens?


Certainly merging is underrepresented in the sample, but by how much I'm not exactly sure.

Quote:
To play devil's advocate, what if it turned out that the no-merge variant was simply deeper and more strategic, even if you see fewer very large bugs in it?


If that were the case I'd be wholly and enthusiastically in favor of no-merge variant! And as I say, I can't be sure it's not the case. So I remain very open to being wrong.

Quote:
Ultimately I guess only testing can decide what works best. Maybe some AI implementations can help. It would at least be interesting to see statistics on how often merges actually happen in a big archive of games. I tentatively hypothesize now (contrary to my initial hasty guess in the OP) that merging may be an unnecessary infrequently used option, and the game is just as good without merging allowed. But I dunno.


It would be wonderful if you're right, because no-merge is less confusing. There's a guy making an app for the game right now, and has a functional (if poor) AI working. I'll see if i can get him to run some tests.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nigel N
Hong Kong
flag msg tools
Avatar
mbmbmbmbmb
My girlfriend and I have only played three games so far, so we are definitely not as experienced. But we also found that the merging groups being allowed in placement but not in growth confusing.

Accidentally came across this post, we will try to play a few games (if she agrees) without merging at all to see how things go.

And thanks for this wonderful little (by size) abstract!
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
Anna and I played another game (testing with no mergers allowed) this morning, on hexhex4, and the biggest group has 7 stones! It was a dramatic game with (necessarily, to get a size 7 group) some big captures. Here white wins on their turn with no legal placement:


Black still has one move left (grow the lower left size 6 group in one of two possible ways).
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
russ wrote:
Anna and I played another game (testing with no mergers allowed) this morning, on hexhex4, and the biggest group has 7 stones! It was a dramatic game with (necessarily, to get a size 7 group) some big captures. Here white wins on their turn with no legal placement:


Black still has one move left (grow the lower left size 6 group in one of two possible ways).


Nice!

This thread has been making me rethink my stance about the rule in question. My suspicion of it hasn't changed, but it occurred to me that maybe, in the presence of significant uncertainty, it would be wiser to give the less-confusing rule the benefit of the doubt, to take an innocent-until-proven-guilty stance toward it. I'm teetering here.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
FWIW we've been playing the no-merge version exclusively (at least one game every day) and it seems great to us. E.g. we just played a hexhex5 which was amusingly dramatic in the end, as it looked like Anna was about to win, but I managed to unexpectedly copy a weird size-5 bug by building a bug between it and two of her big twisty size-6 bugs, thus eating it and making her have more moves. That gave me time to start building a straight-3 in an alley between 2 of her straight-3 bugs and eating them. In the end we I won, in a position from which neither of us had a legal move (so as close as possible in the end...)

An ASCII-art photo of the end:

O . * O .
* * O * * O
O * O * . O *
O * O * . O * O
O . O O * O * O O
O * O . O * O *
O * . O . O *
. * * * * O
O O O O .

5 differently shaped size-6 bugs in the end, plus various 5, 4, 2, 1 bugs...
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nigel N
Hong Kong
flag msg tools
Avatar
mbmbmbmbmb
Hi Russ, can you clarify whether by "no-merge" you mean that you can still place a stone adjacent to your existing bug, unless this placement will merge two existing groups? Or you can't place a stone adjacent to any of your bug at all?

Sorry I don't know how to put my question into words, I hope you understand it.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
nigelreg wrote:
Hi Russ, can you clarify whether by "no-merge" you mean that you can still place a stone adjacent to your existing bug, unless this placement will merge two existing groups?

Exactly. I.e. the current rules as written, except that the placement cannot merge 2 groups. It can increase the size of an existing group as usual. So no merging happens at all in this variant, as opposed to the current rules as written where merging can happen in placement but not in growth.

(For us, this "variant" is now the "real" game.)
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
russ wrote:
nigelreg wrote:
Hi Russ, can you clarify whether by "no-merge" you mean that you can still place a stone adjacent to your existing bug, unless this placement will merge two existing groups?

Exactly. I.e. the current rules as written, except that the placement cannot merge 2 groups. It can increase the size of an existing group as usual. So no merging happens at all in this variant, as opposed to the current rules as written where merging can happen in placement but not in growth.

(For us, this "variant" is now the "real" game.)


Follow-up question: is there anything you like about the no-merge rule beyond the fact that it's more intuitive?
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
milomilo122 wrote:
Follow-up question: is there anything you like about the no-merge rule beyond the fact that it's more intuitive?

Besides the already somewhat compelling intuitiveness (in the sense of merging being inconsistent with the theme), the other big obvious benefit IMHO is that it makes the game's rules more simple and consistent (i.e. by both placement and growth forbidding mergers, instead of one allowing and one forbidding).

In terms of gameplay I notice only subtle differences. In the early part of a game, we didn't seem to actually merge all that often anyway. (I'd be curious to see stats about frequency of mergers by repeated AI plays...)

Here's an interesting endgame difference: sometimes the endgame can be sped up with this variant, since you can drop some singletons in a row with a space between each pair: then you run out of moves quicker. I think I like this, since otherwise there'd be additional turns placing more stones to connect those singletons; possibly the final game outcome would be different, depending on the specific situation, but sometimes it works out equivalently, and either way in these situations the no-merge variant means that there are fewer final "rote/obvious" fill-in moves of the sort which happen e.g. in a cold endgame of Cross.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nigel N
Hong Kong
flag msg tools
Avatar
mbmbmbmbmb
russ wrote:
the no-merge variant means that there are fewer final "rote/obvious" fill-in moves of the sort which happen e.g. in a cold endgame of Cross.

Slightly off-topic, this is exactly the reason we find Cross dull. We feel like we are mechanically filling up the board to avoid loss. Hope this variant would benefit Bug by preventing similar situations, or at least speeding up.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
nigelreg wrote:
russ wrote:
the no-merge variant means that there are fewer final "rote/obvious" fill-in moves of the sort which happen e.g. in a cold endgame of Cross.

Slightly off-topic, this is exactly the reason we find Cross dull. We feel like we are mechanically filling up the board to avoid loss. Hope this variant would benefit Bug by preventing similar situations, or at least speeding up.

To be clear, the situation in Bug-with-merging would only last a few turns at most, I think, nothing at all like the potentially several dozens of cold moves in some Cross endgames.

Here is an example from our most recent hexhex4 game:

O . * O
* * O * O
O * O . O O
* * . O * O O
O O * O * *
* * . O *
O O * O


Here it was O's turn, and I (O) won -- no more moves. Under our no-merge play, this was a closer (so arguably more dramatic) finish: had it been *'s turn, then * would have won due to no moves! But under the rules with merge, * still has a move (merging the lower 3-group with the lower 1-group).

Tangentially, note that we made two groups with 6 stones on this hexhex4 board. I.e. forbidding merges does not prevent larger groups from forming, at least in our games.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
Anecdotal info: we played three hexhex3 games this morning. In only one of them would allowing merging have changed any of our moves, and it would not have changed the winner of the one game, but only delayed the end of the game by 1 turn as the winner would have done a merge in the endgame and then had no move one turn later.

Another of the 3 games was amusingly short, with only 7 stones placed:
https://boardgamegeek.com/article/28355805#28355805
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Craig Duncan
United States
Ithaca
New York
flag msg tools
designer
Avatar
mbmbmbmbmb
I wonder if within the next decade our personal computers will be able to run "omni-AI" programs: we tell the program the rules of the game we've just designed and then -- AlphaGo style -- in a few hours (or even minutes!) your computer has reached a level of mastery of the game that would take humans centuries to reach.

Game designers could use such a tool to gauge how deep their game is, since I presume the tool could be used to estimate an upper range of possible ELO ratings for a given game. (I suppose one could distinguish between uppermost possible ELO and uppermost possible ELO realistically achievable by humans.)

For the current design question in this thread regarding Bug, with omni-AI Nick could compare upper ELOs for Bug-with-placement-merges-allowed verus Bug-with-placement-merges-forbidden. If no difference in ELO, then it'd be easy to forbid placement merges. If by contrast an ELO cost to forbidding merges turns out to exist, then a judgment would be called for as to whether the cleaner, more thematic rule set is worth the cost in depth. That might not be an easy judgment to make. But at least it'd be an informed judgment!

Of course, we can approximate an omni-AI by having lots of organic intelligence (i.e. lots of humans!) play lots of games of Bug with the different rule sets. But a boy can dream...
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
cdunc123 wrote:
I wonder if within the next decade our personal computers will be able to run "omni-AI" programs: we tell the program the rules of the game we've just designed and then -- AlphaGo style -- in a few hours (or even minutes!) your computer has reached a level of mastery of the game that would take humans centuries to reach.

Game designers could use such a tool to gauge how deep their game is, since I presume the tool could be used to estimate what an upper range of possible ELO ratings for a given game. (I suppose one could distinguish between uppermost possible ELO and uppermost possible ELO realistically achievable by humans.)

For the current design question in this thread regarding Bug, with omni-AI Nick could compare upper ELOs for Bug-with-placement-merges-allowed verus Bug-with-placement-merges-forbidden. If no difference in ELO, then it'd be easy to forbid placement merges. If by contrast an ELO cost to forbidding merge turns out to exist, then a judgment would be called for as to whether the cleaner, more thematic rule set is worth the cost in depth. That might not be an easy judgment to make. But at least it'd be an informed judgment!

Of course, we can approximate an omni-AI by having lots of organic intelligence (i.e. lots of humans!) play lots of games of Bug with the different rule sets. But a boy can dream...


I would kill for such a thing, specifically to resolve questions like this. As this is only something that I would expect to matter at high-level play, there's really no way for me to assess my worries at present.

(however, due to this thread, I'm leaning towards going back to the no-merge rule, for the reason I mentioned above)
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Craig Duncan
United States
Ithaca
New York
flag msg tools
designer
Avatar
mbmbmbmbmb
milomilo122 wrote:
(however, due to this thread, I'm leaning towards going back to the no-merge rule, for the reason I mentioned above)

Although I have not yet tried Bug (soon I will!), given the evidence from Russ, the no-merges rule doesn't seem to have much effect on game play compared to the merges-allowed rule.

I'm a bit surprised that merge vs no-merge seems to have such little effect on game play, but that surprise likely merely reflects my total lack of experience with the game.

So in the interest of an easier-to-grasp rule set, you seem to me (FWIW) in the quoted text to be heading in the right direction.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
cdunc123 wrote:
milomilo122 wrote:
(however, due to this thread, I'm leaning towards going back to the no-merge rule, for the reason I mentioned above)

Although I have not yet tried Bug (soon I will!), given the evidence from Russ, the no-merges rule doesn't seem to have much effect on game play compared to the merges-allowed rule.


Again, the use-case I'm thinking of is a much higher level of play than anyone now plays at. Russ' experiences don't relate to that use-case in any straightforward way. The question isn't as cut-and-dried as Russ' experiences (and mine, for that matter, which are the same) would suggest.

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Craig Duncan
United States
Ithaca
New York
flag msg tools
designer
Avatar
mbmbmbmbmb
milomilo122 wrote:
cdunc123 wrote:
milomilo122 wrote:
(however, due to this thread, I'm leaning towards going back to the no-merge rule, for the reason I mentioned above)

Although I have not yet tried Bug (soon I will!), given the evidence from Russ, the no-merges rule doesn't seem to have much effect on game play compared to the merges-allowed rule.


Again, the use-case I'm thinking of is a much higher level of play than anyone now plays at. Russ' experiences don't relate to that use-case in any straightforward way. The question isn't as cut-and-dried as Russ' experiences (and mine, for that matter, which are the same) would suggest.

Right. Good point. (But Russ and Anna may reach that high level eventually... )

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Russ Williams
Poland
Wrocław
Dolny Śląsk
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
milomilo122 wrote:
Again, the use-case I'm thinking of is a much higher level of play than anyone now plays at. Russ' experiences don't relate to that use-case in any straightforward way. The question isn't as cut-and-dried as Russ' experiences (and mine, for that matter, which are the same) would suggest.

Do you have any specific concrete examples of play where you think that allowing merging could cause deeper more interesting play for top-level players?

I'm sincerely curious, because (based on my admittedly limited experience so far) I confess that I don't see any strong reason to suppose that it would make a significant difference for the depth or interest of top-level play, or to suppose that allowing merging would be better for top-level play.

I.e. in an alternate universe, alternate-Nick published Bug without merging allowed, and alternate-Russ proposes allowing merging, and alternate-Nick hesitates in the same way, due to worrying that this would hurt high-level play...


(And then in that alternate universe, alternate-Russ asks "Why? It's not clear to me one way or the other. I.e. in an alternate universe, the game was web-published with merging allowed, and Russ proposes disallowing merging, and Nick is uncertain...")
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nick Bentley
United States
Madison
Wisconsin
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
After playing a bunch of games this weekend where merging *was* important, fun, and allowed groups to get bigger, I'm now squarely back on the fence (I didn't record any of them, unfortunately, having played in quick succession).

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
George Leach
United Kingdom
Godalming
Surrey
flag msg tools
designer
Avatar
When I played I obviously misread and thought that merging was not allowed through both growth or placement.

I think it obvious that merging in placement would allow a slight increase in depth since there are choices available that wouldn't otherwise be available. I suppose we can't be sure that those options don't also allow for a collapsing of the tree too...
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
1 , 2  Next »   | 
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.