The current switch implementation was changed (several times I might add) because a number of people where unhappy with the other algorithms I came up with.
I remember some of these discussions on our chatserver. I hope we have here a nice thread where anyone can participate. Actually, I think that almost anyone is unaware how the algorithm works now. You see my guess in the previous post, I have just figured out that it works quite differently. The unoccupied planet closest to the sun is
ALWAYS (whether you have min or not) changed by the minstart switch in the following way:
Code: Select all
UP T lg --- P M ng
UP S lg --- P M ng
UP M lg --- P M ng
UP L ng --- P L ng
UP H ng --- P H ng
P T lg --- A M ng
P S lg --- A M ng
P M ng --- A M ng
P L ng --- A L ng
P H ng --- A H ng
A T lg --- R M ng
A S ng --- R M ng
A M ng --- R M ng
A L ng --- R L ng
A H hg --- R H ng
R T ng --- UR M ng
R S ng --- UR M ng
R M ng --- UR M ng
R L hg --- U L ng
R H hg --- U H ng
UR T ng --- P M ng
UR S ng --- P M ng
UR M hg --- P M ng
UR L hg --- P L ng
UR H hg --- P H ng
I was quite surprised to see this result. It was several times mentioned here that UR ngs are a serious problem for many players. In almost any ladder match they are restart (a popular 1v1 (esp. Ladder) restart rule seems to be: 1 urng or 2 rng/urhg) and even in 4ways some players propose a nourng rule (see Steps motd). The drawback of the current algorithm is the huge increase of UR ng or R ng planets. So far I thought this problem was caused by the popular planets=4 switch but now I see it is actually caused by /minstart. I observed a sample of 100 minstarts and received following results:
Without minstart I received (just the unoccupied planet closest to the sun):
10 rich small or medium ng
2 urich small ng
By using the minstart switch we receive (same maps!):
46 rich medium-huge ng
14 urich medium-huge ng
So we have now lots of restarts when we use the above-mentioned rules or we have 4ways where some uni tol has a huge advantage (imho pointless to play such maps). More restarts or imbalance. I really think that the switch causes more problems than it actually solves.
Because one of the older algorithms would only modify the system if it failed to meet the old TSL rules (although I can't find the rule now...seems to be removed now
I still thought that it works this way. And I really would like to see that the switch works again this way...ok...we have a problem that we have to define the rule
I propose Step and/or Gusset (if they want) should moderate several rules threads where we discuss extensions (or changes) of the old rules.
Further, when we use the switch then it should IMO work like:
If up then poor med
If poor tiny or small then poor med
If abundant tiny, small or huge then poor med
If rich huge, large, small, tiny then abundant med
If ur then rich med
In detail:
Code: Select all
UP T lg --- P M ng
UP S lg --- P M ng
UP M lg --- P M ng
UP L ng --- P M ng
UP H ng --- P M ng
P T lg --- P M ng
P S lg --- P M ng
P M ng --- P M ng
P L ng --- P L ng
P H ng --- P H ng
A T lg --- P M ng
A S ng --- P M ng
A M ng --- A M ng
A L ng --- A L ng
A H hg --- P M ng
R T ng --- A M ng
R S ng --- A M ng
R M ng --- R M ng
R L hg --- A M ng
R H hg --- A M ng
UR T ng --- R M ng
UR S ng --- R M ng
UR M hg --- R M ng
UR L hg --- R M ng
UR H hg --- R M ng
In combination with planets=4 or planets=5 the minstart switch should fix the map quite rarely.
Using p=4 I had 7/10 TSL minstarts without switch. And with p=5 I had 9/10. And the missing ng planet was compensated by a large urhg. Surely a playable start. So I prefer at moment not to use minstart.
Many people do not like any Heavy planets in their home systems.
In most cases I don't like abundant hg or rich hg either (when I want to play a techrace it is ok for me to have some since I don't settle them early and my prodrace opponent might have some hgs too). But the minstart switch shouldn't fix this problem. And actually it doesn't solve it right now (it just changes one planet). So I would propose a medstart switch where you use the proposed algorithm always for all planets. (Of course different algorithms are also possible....alex will soon propose some further ideas.)
Many people do not like good Normal G planets in their opponents home systems.
Currently, the minstart switch improves uni tol far too much.
So the above planets were not generated and we still had many many restarts
But this was before Cybersaber retired? Right?!
When someone complains about a bad start he should use the richstart or goodstart (and hugestart). Richstart, hugestart is Inver’s preference at moment AFAIK. Surely more fair than the usual minstart gambling right now. ( I prefer to play p=5 without further hw switches (4ways). p=4 without minstart should be ok in 1v1s.)