SGOTM 13 - One Short Straw

I don't think throwing up our arms is the way to go. I'm away from the game right now but feel that the SIP capital is better long term, especially with 4 food to itself.
 
Five teams have settled on PH, it seems.
Which is an uninspiring statistic. We don't know how many of them ran any tests or to what extent they did so.

Similarly, any team who decided to settle in-place gives us no additional info.


So either PD went somewhere no one else went and liked it, or they saw something they liked that no one else liked enough (others may have preferred to stay on the PH), or PD decided to come back to SIP. Very bizarre. Maybe on epic they liked SIP best but didn't want to risk missing something really good.
It's this info that is the most interesting.

Out of those proposed scenarios, the last one, about liking SIP while being willing to spend a few Epic turns, sounds the most likely.


However, we could probably sit here and come up with other scenarios. What if most teams sent their Warrior to the north of the PH, while PD sent their Warrior SE of the PH? Maybe PD found something, like a Grassland Gem SE + S of the "3E" location, and decided to settle 3E with their capital.

The opposite scenario for where everyone sent their Warrior is unlikely (I can imagine most people wanting to scout out the potential for extra land to the north if they did not do perfect fog-gazing), but is still possible, meaning that PD could have even moved to the north to settle.

Still, given how I imagine that most teams who move to the PH will send their Warrior northward first, I find it very unlikely that PD moved their Settler further to the north and thus to me, it seems most likely that PD either moved their Settler to 3E or back to in-place.


To me, the data tell me that SIP is better, simply because they're close enough that the increased research rate plus SIP being more powerful in the mid-game rapidly moves it ahead of PH in hammers.
SIP is very strong on research, while PH gets us Work Boats out early.

The problem with PH is that "there is nothing else to build" for so long due to being slow on Research. FORTUNATELY, we actually want some exploring Work Boats--if we did not, then SIP would be the clear winner.

I really can see the effect of fast Research helping out a lot. SIP just seems to get so much better timing as a result of getting buildings and crucial techs like Bronze Working for Slavery so much sooner. I am starting to get sold on this whole "he who researches fastest, wins fastest" idea.

The cost of that additional research, though, comes at the cost of slower or less exploration. SIP CAN push for faster Work Boats at the cost of some Research. PH, however, doesn't seem to perform so well doing the opposite, so PH really is limited to a slower Research rate but better exploration.


Right now, I kind of want to see the updated spreadsheet comparison. I wonder if we should email our Turn 82, 1950 BC saves to Mitchum and then see if he can post the updated screenshot on a free image-upload site? No worries, I can also patiently wait.


At present, though, I am leaning towards a SIP option where we plan to commit to one permanent early exploring WB that comes before Settler 2 (as per the only vote on the subject) and one temporary exploring WB that soon-there-after gets netted. This approach gives us an early exploring WB that can head off almost directly to go and meet the AIs, while City 2's location is already known (3E). Ultimately, I think that we'll meet the AIs the fastest this way, as:

a) We won't have to mess around trying to find a location for City 2 and by the time that we're ready to settle City 3, we'll have had some partial exploration from an additional WB and possibly from our Galley. In this way, our first WB explorer can go directly to the AIs ASAP

AND

b) We can get this first WB explorer out FASTER using SIP than PH, since we can tech to Bronze Working (and hence, Slavery) faster using SIP. Ultimately, we will meet the AIs the fastest by a few turns this way


Also, SIP gives us access to early Sailing and Writing, meaning that meeting the AIs early actually has a purpose. Until we get these two techs, meeting an AI has relatively limited value pre-Alphabet, but once we get them both, our research rate will further take off, pulling ahead of PH in terms of Research at an even faster rate.
 
Okay, guys. Shyuhe requested that we make our decision this weekend. I respect our captain's requests. I vote for PH. It's a strong opening. It gives us good exploration and a balanced start. It's relatively simple, although the simplest opening is not the best. Better is working the clams for 5t to get fishing-mining on T20, the worker built on T19 and the worker improves mine1-cornfarm-mine2, as best I can determine.
 
It looks like the PH wins...

@LC I'm surprised that you suggest that we build the mine before farming the corn. Why is that? I would think growth trumps hammers. Are you doing it to speed up WBs? Does it pay off in the long run by forcing us to work clams for the extra commerce?

Rather than just focusing on the first 20 turns, where does doing this leave us at T82 so we can compare to the other games? I assume that this was one of your tests that you have not been able to upload...
 
Well, I still stand by SIP, in that we:

a) can get out the fastest Work Boat Explorer #1, faster than any PH test run posted to date

AND

b) get better Research than all of the PH test runs

AND

c) have Settler 3 on Galley 1 at the average time time as PH gets Settler 3 on Galley 1, i.e. Turn 85

AND

d) have the flexibility of Chopping a Forest that was not chopped in order to turn Work Boat Explorer #2 into an earlier Galley 1, if we choose to settle City 2 off-island instead of settling it 3E*

AND

e) have good infrastructure set up in that we have a Lighthouse and a Granary


* Note that I have not tested this approach, but the Worker around this time is just building Roads, so it could instead be chopping the 2nd Forest on our initial island, and 45 Hammers for a Work Boat + 30 Hammers from a Forest Chop = 75 Hammers for a Galley.

We can get all of this with my saved games from message #247. Unfortunately, we haven't been able to see the updated spreadsheet which showcases this test run, but I think that this game shows that SIP can explore just as well as PH but can do it better (getting a Work Boat out a few turns faster than any PH run), while still giving us superior Research (which in combination with the early Work Boat gives us the fastest Foreign Trade Routes), good infrastructure, and 6 Health in the capital.


Yes, some SIP build orders and tech orders don't give good results. The same can be said for several PH build orders and tech orders. But, we have one for SIP that seems to meet all of our goals and does the two that we seem to care about the most (early exploration and a good research rate) and does them better than any PH test run that I have seen. So, while on average, SIP may not beat PH, this particular SIP test run does beat all of the PHs that I have seen given our preferred criteria, and we will only play out one single test run, not the average of multiple test runs, so all we care about is the best test run, given certain criteria.

Given my understanding of the team's desired criteria, SIP from message #247 is the clear winner.

But, if you're going to claim that PH is good, WHICH TEST RUN from PH do you like the best? Because we can ONLY RUN ONE PATH in the real game. I don't know about you, but I'd rather take the single path that best meets our goals, instead of picking a settling location where the AVERAGE path best meets our goals.


It looks like the PH wins...
It was my understanding that people were waiting to see the final test result comparisons, since time was allowed for additional testing. There's no point in having run additional testing if the additional testing is not allowed to sway anyone's vote, correct?


As far as I am concerned, I am happy with SIP from message #247. If people aren't convinced, then it means re-examining our judgement criteria and figuring out what those criteria are and picking a test run based on this new information.
 
@Dhoom

I tend to agree that SIP is better; having played all three starts (Corn, PH and SIP), SIP seems to be a more viable long-term capital, even moreso if we skip 3E and Paris gets 4 food tiles. I said "PH wins" because several team members don't seem to be swayed by any of the testing going on and one SIP proponent just flipped... :mischief:

As a team, it would be nice to have players more involved and egaged. It's tough to discuss/debate the issues when someone makes a simple statement like: "I don't like SIP" or "I prefer SIP." I think we should explain WHY we are for or against a particular action. Then, as a team, we can come up with solutions that address everyone's ideas.

As an example, many pro-PHers said that they prefer PH because they put a higher value on early exploration than research. Dhoomstriker then showed that we could have the best of both worlds with early exploration AND faster research. LC showed that SIP has as much or more food-hammers than PH.

So I guess I'd like to hear an argument for why PH is better than a SIP save that gets similar early exploration to a PH start. I'm all for moving the game forward, but I think it's okay to delay another day to get everyone's reasons on the table.

BTW: I know I seem wishy-washy on SIP vs. PH. I've always preferred SIP but was willing to go PH to get the game going as the majority wanted and to keep everyone's interest level up. In addition, PH isn't a bad start so I figured we could still do well with it...
 
Been remaining quiet on the settling aspect as I haven't been able to run enough tests and I can't seem to decide just from reading the posts. It seems very close to me and I'm fine either way. There are too many uncertainties about the map layout for me. Testing this thoroughly for 82 turns when you only have turn 0 intel is extreme.

Mitchum said:
So I guess I'd like to hear an argument for why PH is better than a SIP save that gets similar early exploration to a PH start. I'm all for moving the game forward, but I think it's okay to delay another day to get everyone's reasons on the table.
Sounds like SIP is best then. The only minor argument I can come up with is that a plains hill is a poor tile to work and that we get extra output per poulation point throughout the game (moreso noticed at lower sizes).

If SIP has just as good research, exploration and production I see no reason not to SIP. ;)

But my main wish is to get the game going. Seeing as this is so close there will be other decisions to make later that will have a bigger impact.
 
Dhoom, your post 247 and the saves you posted don't seem to correspond. The saves only show 3 wbs built, 2 netted, 1 exploring. Your post says you have 2 exploring wbs.

Otherwise, I don't find your T84 save to be comparable to mdy's and my best PH efforts (not yet uploaded), in which both have our second city with a granary circa T82-83. In other words, you have 1 less wb and 1 less granary, in exchange for more research and +1:health:.

In my SIP, I only have one explorer, but earlier than mdy's (28 explorer-turns vs. 30 exp-t), but I have all three clams netted, so I have one more wb than you and I settled Orleans offshore, and my Orleans will finish the granary 3-4t sooner than yours.

---------

In any case, to make a good comparison, imo, you would need to settle the second SIP city offshore, because onshore looks like a serious drag on Paris.

I still can't upload files, so I'll email them to Mitch.
 
@LC I'm surprised that you suggest that we build the mine before farming the corn. Why is that? I would think growth trumps hammers. Are you doing it to speed up WBs? Does it pay off in the long run by forcing us to work clams for the extra commerce?

Rather than just focusing on the first 20 turns, where does doing this leave us at T82 so we can compare to the other games? I assume that this was one of your tests that you have not been able to upload...
In my testing, PH has two early problems: slow research and no hammer tile for a quick wb. Mining the hill gets the wb faster to speed up research, plus it gives us an extra 10b in the process, working the clams for 5t to slow down the wkr but speed up mining. (It's an elegant solution--try it!) The clam nets are growth and commerce, let's not forget, so it's not exactly a question of growth versus hammers. It's hammers producing growth (and commerce). Furthermore, I can't find any other way to get Pottery by T53 which gets us an earlier granary (1t), which again produces more growth (1t).

I haven't tried it with sailing first yet, to compare it to mdy's save. That would be the real test. That's something we need to do no matter what, because we need to decide which is stronger, sailing (REX) or pottery (growth) first.
 
Back now, but have quite some catching up to do on test saves.

Agree w Rusten. I'm mainly concerned with getting this rolling, as we'll have plenty of other major decisions to make and lots of minor ones when the warring gets going. It seems that you guys have proven, if anything, that the two spots are fairly well balanced.
 
I would also like to get this rolling and am ok with either spot. I'm going to wait until the end of the day to see if any of the other PH'ers chime in with anything. I think the current count looks something like:

shy: abstain
Mitchum: PH
Dhoom: SIP
bbp: abstain
Ras: ---
mdy: PH
LC: PH
EP: PH(?)
ungy: PH(?)
Rusten: abstain
Lexad: MIA

I've changed anyone who's expressed ambivalence to abstain. So yes, the two spots are quite comparable in the eyes of the team. Given the time table we have to work with, I'd like to move forward with something by the end of today...
 
PhotoBucket to the rescue. I got LC and Dhoom's latest saves. I've added those 4 plus my latest to the table. New entries are highlighted green.

It looks like LC #257 and Mit #250 both tried a similar approach of getting a granary up in Orleans ASAP. LC can whip it next turn while mine finished a few turns ago. Of course, none of this matters too much because we will likely find a better place to settle in the real game.

Dhoom #269 is similar to LC #257 in that they both have Settler #3 on a galley already by T81, the fastest we've seen so far.

@ Rusten

I agree that testing to T82 seems extreme and our real game will likely diverge from the test game ~T50 once we start sending out explorers. The reason that we played tests this far was to see how the different capital locations would develop as the second and third settlers were being produced. It's too hard to tell the difference if we only run tests to T50 or so.

Spoiler :
 
I've changed anyone who's expressed ambivalence to abstain. So yes, the two spots are quite comparable in the eyes of the team. Given the time table we have to work with, I'd like to move forward with something by the end of today...

Keep in mind that once we settle on SIP/PH, we also have to agree on which test run we want (or at least the first XX turns or so). That may or may not be an easy thing to do.

BTW, how long should this first set be? My guess is that we could play until BW is in. There really isn't much going on before then. No exploring yet. PH has a worker, so we'd have to agree on which improvements to make first. Playing this far would let us see where copper is. If it's on our island, it could change our MM...
 
Once we start playing, we will want to take screenshots of the F8 demo screen each turn so that we can try to glean additional information about the AI. Other than that though, yes I don't think there's anything much going on until a worker/WB is completed.
 
Looking at the saves and the data, my current preference is PH - LC #257

I prefer this save for 7 reasons.

1. It already has settler #3 out and will quickly grow back to 5 pops.
2. It traded an early City #2 for an earlier Pottery/granary. This must be why the capital is in good shape after just having whipped a settler.
3. Whip unhappiness isn't too bad in Paris.
4. A later city #2 could provide a slight advantage. This save has 28 exploration turns before we need to indentify where to settle.
5. Two early wb explorers (vs. most SIP test). As has been said, having 2 explorers, each headed in a different direction, could end up having a huge impact. With only one WB, if it heads the wrong direction, it could be a long time before we meet anyone.
6. SIP/3E seems stong, but long term, 3E stealing food from Paris could become an issue.
7. I hate to play the resource guessing game because it can lead to bad choices, but if there is a strategic resource on the island, it will be in PH's BFC and can be worked. With SIP, the two grassland hills are outside the BFC. This is a very minor issue to me though.

BTW, LC #257 did NOT use the Fish as far as I can tell. That cell was RED because of how Excel formats columns when inserting a new one. Any save using the Fish was ignored in my evaluation because using it, even if uniproved, does have an impact on the outcome.

Can others propose their favorite test run? With that info, we should be able to move forward and actually know HOW to play the first 30 turns, not just where to settle...
 
Dhoom, your post 247 and the saves you posted don't seem to correspond. The saves only show 3 wbs built, 2 netted, 1 exploring. Your post says you have 2 exploring wbs.
Weird. Perhaps you are looking at the wrong saved game file? I just downloaded the 1950 BC file from message #247.

I see two netted Clams and 2 Fortified Work Boats. We are working 3 Clam squares (each of them Lighthouse-enhanced), so we certainly could have benefited from settling a 3rd Work Boat, but in the interest of demonstrating that "SIP can explore, too," I dedicated two Work Boats to exploration.

Looking at Mitchum's table, that gives us the earliest Work Boat Explorer #1 out of any recorded test runs. Work Boat Explorer #2 comes later than most PH games, but the amount of "Exploring turns (WB only) by T82" seems to balance out to a good total amount (greater than 50 turns).


Otherwise, I don't find your T84 save to be comparable to mdy's and my best PH efforts (not yet uploaded)
Kindly find a way to get them to Mitchum so that we can all see and compare the results, please.


in which both have our second city with a granary circa T82-83. In other words, you have 1 less wb and 1 less granary, in exchange for more research and +1:health:.
Well, it sounds like it is only 1 less Granary, since I do have 4 Work Boats. As noted, I dumped City #2's Hammers into a Galley, in order to save on whipping unhappiness in the capital.


In any case, to make a good comparison, imo, you would need to settle the second SIP city offshore, because onshore looks like a serious drag on Paris.
I'm not really seeing the major difference here.

Let's see... if we settle on the island to the west, we will steal one of Paris' Clams. That leaves Paris with 3 Food Resources.


In fact, I'd rather let a second City steal the 5-Food Corn than a second City steal the 5-Food-2-Commerce Clam, since our capital is likely to get an earlier Library (and hence make better use out of the Commerce in the short run) and will almost certainly later use Bureaucracy (and hence make better use out of the Commerce in the long run).


Settling 3E can easily make do with just the Corn, leaving Paris with 3 Food Resources. That allows 3E to work: its City Centre (2, 1, 1) its Corn (5, 0, 0), its GH Mine (1, 3, 0), multiple Coast squares (2, 0, 2 each), and 2 Scientists (-2, 0, 0 and 3 Flasks and 3 GPP each) at its happiness cap.

For a City with 1 less total City Maintenance at City Size 1 than PH + settling off-continent on the western island, we have a decent, low-maintenance City that *allows our Work Boats to meet the AIs for Foreign Intercontinental Trade Routes ASAP,* instead of having our earliest Work Boat Explorer be forced to scout for City #2's location. Given that we also have a better research rate than any PH run, it means that these 2 Cities will have the best possible Trade Routes faster than any PH run.


In my SIP, I only have one explorer, but earlier than mdy's (28 explorer-turns vs. 30 exp-t), but I have all three clams netted, so I have one more wb than you and I settled Orleans offshore, and my Orleans will finish the granary 3-4t sooner than yours.
Since you claim to have 4 Work Boats and since I double-checked Dh #247 to ensure that it has 4 Work Boats, we have the same number of Work Boats.

I am assuming that you are referring to LC #107, which has less exploration, a bit more
Research, a similar date for City #3, and according to your claim, a bit less production in City 2.

We know that there is a bit of trade-off in production and research in order to get that extra exploration for SIP. The team seems to value that extra exploration highly. What compels me about Dh #247 is that we get the amount of exploration that the team seems to want, a level of exploration that is easier to achieve with PH but was also achieved with this SIP test, while still maintaining more Research than all of the PH runs.


mdy's and my best PH efforts (not yet uploaded), in which both have our second city with a granary circa T82-83
I am very interested to see how they will stack up against Mit #250. Can you improve upon Mitchum's result?

Also, if we value production very highly and don't really care about research, then Dh #269 is a compelling PH game. Throwing research to the wind, we've already started on Settler #4 when Turn 82 rolls around.
 
@ Dhoom

If you check the latest table in post 293, you'll see the details for LC's latest PH save, which is #257.

EDIT: Don't forget that the PH saves have an idle galley, which can explore the nearby islands. This "free" exploration is NOT included in any of my numbers.
 
Looking at the saves and the data, my current preference is PH - LC #257
There are multiple test runs mentioned in that message. Can you please clarify which one of those test runs it matches up to?

Also, did we use a Forest Chop on the western island in order to get the Granary? If so, I don't think that it's a fair comparison, because my understanding was that we were supposed to settle on the Grassland Forest square when settling on the western island. If we got up to 45 Hammers without a Forest Chop, then that's a decent amount of production for City #2.


1. It already has settler #3 out and will quickly grow back to 5 pops.
Yes, that's cool. This production comes at a cost of some research and exploration, but it is what it is.

4 turns saved on City 3 compared to Dh #247 at the cost of more than 50 Flasks and a slower Work Boat Explorer #1 (11 turns slower).


2. It traded an early City #2 for an earlier Pottery/granary. This must be why the capital is in good shape after just having whipped a settler.
City 2 comes at the same time as Dh #247, so that's fine.


3. Whip unhappiness isn't too bad in Paris.
Quite compariable to Dh #247.


4. A later city #2 could provide a slight advantage. This save has 28 exploration turns before we need to indentify where to settle.
Same for Dh #247, except that if we settle 3E in Dh #247 instead of trying to chop an earlier Galley as I suggested might be possible, then we get even more time to identify a settling location, otherwise these saves are equal on this point.


5. Two early wb explorers (vs. most SIP test).
Equal to Dh #247, except that Dh #247 has the first Work Boat Explorer for 11 extra turns. If we use our first Work Boat to meet the AIs, that's potentially 11 extra turns of Foreign Intercontinental Trade Routes, depending upon how far away the AIs are relative to when we learn Writing.


6. SIP/3E seems stong, but long term, 3E stealing food from Paris could become an issue.
I think I just showed in my last message that it isn't a major issue at all, since we'd only need to steal the Corn in the long run and perhaps in the short-run, too (only borrowing a Clam temporarily if Paris is whipped down to Size 2 and can't actually use all 3 of them at once).


7. I hate to play the resource guessing game because it can lead to bad choices, but if there is a strategic resource on the island, it will be in PH's BFC and can be worked. With SIP, the two grassland hills are outside the BFC. This is a very minor issue to me though.
As I showed, 3E works out quite well as a City. If 3E is settled on a Strategic Resource, we will still get +1 Hammer for it, unlike if 3E were a flatland Grassland square, in which case we WOULD miss out. So, with 3E being a GH square, 3E would pick up any benefit from such a potential Strategic Resource quite nicely.
 
Let's compare LC #257 (PH) to Dh #247 (SIP):

1. LC #257 did use the forest chop. I don't recall forcing a settling location for the island to the west. The only constraint was that the Fish tile couldn't be used. LC, did chop the forest into the granary, as did many PH starts. I don't see an issue with this. It shows that the worker has spare turns and can be available to make improvements in City #2, even if it is settled circa T70. IIRC, with SIP, the worker will still be busy around Paris since it comes a bit later. So, if city #2 is off continent and not 3E, sending the worker with the settler could prove problematic.

2. LC #257 already has a galley and already has settler #3 out. Dh #247 still has to whip the galley and settler, moving Paris back to 2 pops and Orleans back to 1 pop. When LC #257 does crack the whip in Orleans next turn (or there about), it will be to complete a granary, not a galley. So, in a sense, LC #257 is ahead a granary in Orleans and a lot of hammers in Paris (settler #3 already whipped). Paris is also, in effect, larger since it has already whipped the settler and is growing back to pop 4/5. Dh #247 Paris will be whipped back to 2 pops shortly and thus be behind about 2 pops.

3. Dh #247 does have +1:health: in Paris and 2 more potential chops. This is production in reserve that LC #257 doesn't have.

4. Exploration is about a wash. Dh #247 gets WB#1 out sooner but WB#2 out later. LC #257 does get some "free" exploration with the galley that Dh #247 does not.

5. Dh #247 is ahead 53 beakers, or about 3 turns of research.

6. In LC #257, Paris has already whipped settler #3. Dh #247 has not, so we can effectively add +15:mad: to Dh #247.

7. Most other points are farily similar (e.g. clams netted, etc.)

If we find a really juicy off-island location, any PH start is ready to take advantage of it since PH starts from the beginning with the idea of needing to settle off-island. For any SIP test that "plans" on settling 3E, some compromises would have to be made to settle that location as city #2.

As many have said, I think a PH start is a bit more balanced as will allow us to better capitalize on things as our exploration progresses. We get this and the advantages listed above at the cost of 3 turns of research...


BTW, I don't know which test run LC #257 corresponds to, but I got the save from LC. He'll have to fill us in on that.
 
Top Bottom