Civs from easiest to hardest

Arathorn

Catan player
Joined
Jan 10, 2002
Messages
3,778
Location
Illinois
What civs are easiest/hardest to play? As far as I can determine, the key factors are civ-specific traits and Unique Units (UU). The manual claims that diplomacy is easier with culturally linked neighbors, but I'm not sure how large a role, if any, that plays, so I am ignoring it.

All of the following are based purely on my opinion. I’ve not played all the civs. I’m also assuming a high difficulty level (deity or maybe emperor). I'd be interested in hearing others' views.

Easiest:
Iroquois – Mounted Warriors are a dominant ancient age force and fully upgradeable all the way to cavalry (useful through most of the Industrial Age). Religious is an awesome trait with cheap temples and one-turn anarchy. Expansionist is marginally useful for early map-selling. The UU advantage is simply huge, though. Only "downsides" are an early GA (but that can lead to lots of early wonders and an early lead, which is never surrendered) and the need for horses for the UU.

Aztecs – Some would disagree but their UU is just so awesome. Early retreat, extremely cheap, lots of early leaders make these little guys way cool. Pillaging at speed. Add in militaristic and they promote and crank out leaders. Religious is almost too much to add. The lack of upgrade path for jaguar warriors and the incredibly early GA are the detriments.

Japanese – Samurai are both the best offensive and defensive unit for a reasonable period. They prevent retreat, defend as well as musketmen (their contemporaries), only cost 10 shields more, and attack well. Militaristic and religious (again) for leaders and some happy culture. Start with "The Wheel" so can see horses early. Downside is that you have to wait for Military Tradition to upgrade those horsemen you have running around.

Still Easy:
Egyptians – War Chariots are essentially horsemen without horseback riding. Very early move 2 with reasonable offense and upgrade directly to knights. Religious and industrious work well. Cheap culture/happiness and lots of improvements from fast workers. Downside is the very early GA and a struggle with mountainous/jungle-infested maps.

Zulu – Hate having your horsemen attacked after they do their stuff? Me, too. The answer is the Impi, who is able to keep up. Horsemen don't retreat from these guys, either. Upgradeable all the way to Mobile Infantry. They don't need a resource to build. Their Militaristic nature makes for veteran impi/horsemen hordes and early GLs. Expansionist helps them know where to attack and gives map selling options but is generally weak. Again, an early GA is the downside, as well as the generally weak expansionist trait. Also, culture can be a problem for the Zulus as everything is full price for them and they have no shield bonuses.

Persians – The highest offensive number before knights belongs to the immortal, Persia's UU. Still, I find I like legionaries better. I ranked Persia as 'easier' than Rome, though, for a number of reasons. One, Persia starts only one advance away from iron working, since it starts with bronze working. Industrious workers make getting a road to those enemies a lot easier. Scientific is nice, with both cheap libraries/universities and those three free scientific advances. Downside is that the UU can't retreat, leading to high casualties (although an ancient army of three immortals is pretty invincible). The UU also doesn't upgrade, leading to obsolescence problems.

About average:
Greeks – Hoplite is an awesome defensive unit up until riflemen (cost compare with musketmen and I know which I’d rather have). It upgrades all the way to mech infantry, too, which is nice. Commercial and scientific should lead to huge tech leads, but it doesn’t seem to any more than any other civilization. Starting on the road to the Great Library is nice, but science shouldn’t be the Greeks problem. Downside is no good offense, hoplites are too slow to really escort horsemen, and the early GA has very little going for it.

Chinese – Riders get an extra movement point. Since railroad is still quite a ways off, this gives significantly faster response to invasions and a faster invasion force. The ZOC is nice on paper but is less significant in practice. On the other hand, they fight like knights, I don’t think they can run from knights, and can’t be upgraded to. Industrious and militaristic make a nice combination, as wars run smoothly and infrastructure on the land is quickly and easily replaced. The GA is perfectly timed for all those nice Middle Age wonders (Sistine, Bach, even Leonardo and Sun Tzu). The inability to upgrade horsemen is a big problem, as are the long periods of anarchy. China is often at war and hence will occasionally lag technologically.

Babylonians – Bowmen are a combination spearman/archer. They are a nice shock troop in the ancient age, doing two roles, but their upgrade path stinks. They trigger an early GA which is probably not needed. On the other hand, the twin traits of scientific and religious make the Babylonians the undisputed culture king of Civ3. They have lots of cheap city improvements and should be good at cultural absorption and minimizing flips after conquest. The free tech can be very handy, too. The biggest downside is the adequate but horribly unupgradeable UU. Who needs longbowmen?

Romans – Legionaries can, indeed, rule the Ancient Age. At 3/3, they’re the best defenders around (until musketmen, who aren’t much better and a lot more expensive). They are slow on the attack and no better than swordsmen that way. They also have no upgrade path and require iron. Militaristic is nice for getting elite legionaries and leaders, but legionaries do die a lot more than fast units, since they can’t retreat. Commercial is rarely helpful. The Romans’ other downside is a huge difficulty building cultural improvements, as they have no extra shields and everything is full cost.

Getting Difficult:
Germans – Panzers are nice, the extra move and blitzkrieg ability make them terrors of the late industrial age. On the other hand, modern armor isn’t that far away, they still really struggle against mech infantry, tanks will still crush them, and panzers come too late in the game to really matter. By that point, a game is usually won or lost and panzers rarely make the difference. The GA really comes at a pretty inconvenient time, as few wonders still need to be built and war is a LONG process. Militaristic is OK, but it’s had to completely take advantage of early, when it’s most important. Scientific helps get to tanks a tiny bit faster, but it also serves to shorten the time until modern armor. A very late GA is probably not going to help, the extra movement point isn’t enough with RR everywhere, and a short period for a UU make the Germans kind of tough to win with.

Indians – War Elephants are nice if you have no iron or no horses, but if you’re in that situation, you probably have more problems than just getting knights will solve. Elephants crimp the upgrade path for horsemen, too, for little/no additional merit. At least the Indians are religious, which can solve a lot of problems. Commercial helps some with gold, which will be sorely needed. The GA comes at a great time, for building those Middle Ages wonders, but with no other helps, it’s a tough climb.

French – Musketeers are weird, in that their primarily ability is that you get to keep building a superior older unit (pikemen). For the same shield cost, two pikemen provide much better defense and can upgrade to two good units. Musketeers have better offense than musketmen but are still basically worthless, except to trigger a GA. The French GA is also in a pretty sweet spot for wonders and getting a lead on the AI. Industrious and commercial traits encourage the French to have big cities to get their bonuses. They can also afford a slightly larger empire, with the commercial trait. How they get that empire is a mystery to me.

Hardest:
English – Man-O-War is an updated frigate that has a little better attack/bombard values. They can sometimes sink ironclads, but the age of sail is so short. And Magnetism doesn’t lead to a particularly valuable GA, other than the fact that it comes late. If you need coal, you can leverage your GA into an attack, but that’s not terribly realistic. Expansionist and commercial the British may be, but that is hardly an advantage, as they have no culture benefits, no shield benefits, and a world map only has value so long.

Americans – The F-15 comes extremely late, is better at bombard only than a jet fighter (why not use a bomber?), and requires special resources. The very late GA is probably too little too late to save a player that’s behind and unnecessary for one that’s ahead (one could build the SS faster, I suppose, but tech is usually more important than shields. To get the late GA, you need to start a war and that’s a pain on tech.) Expansionist and industrious isn’t much better than the British situation. The industrious workers and shields help a little, but the poor UU is a lot to overcome. It breaks an upgrade path, too, to add insult.

Russians – The Cossack is the only UU I wish I could simply remove. It breaks the horsemen/knight upgrade path, which is horribly painful. It’s only marginally useful, as defense for a fast offensive unit usually is. You have to build all of them from the ground up, they’re barely better than cavalry, can’t upgrade, and put a GA in an odd time slot. NO THANKS! Add in questionable value in expansionist, and the scientific trait can’t bring them up. The main offensive thread is completely disrupted and knights hang around way too long. It’s a huge hurdle with no reward for leaping it.

Arathorn
 
Originally posted by Arathorn
Germans – Panzers are nice, the extra move and blitzkrieg ability make them terrors of the late industrial age. On the other hand, modern armor isn’t that far away, they still really struggle against mech infantry, tanks will still crush them, and panzers come too late in the game to really matter. By that point, a game is usually won or lost and panzers rarely make the difference. The GA really comes at a pretty inconvenient time, as few wonders still need to be built and war is a LONG process. Militaristic is OK, but it’s had to completely take advantage of early, when it’s most important. Scientific helps get to tanks a tiny bit faster, but it also serves to shorten the time until modern armor. A very late GA is probably not going to help, the extra movement point isn’t enough with RR everywhere, and a short period for a UU make the Germans kind of tough to win with.
[/B]

Alright, take a look at the now over and WON German SG, you will change your mind. The Germans are the best and most fun civilization to play, and the panzers are the best unit, even vs mech infantry. Of course modern war is a long process, that is the fun. Panzers make a HUGE difference. And if you were good enough of a player you would get the GA earlier via wonders. And if you think they are hard, that is the fun in it.

Anyway, I agree that the Americans are pretty hard to play. And the English, but they are pretty fun.
 
I read your ambitious analyses with interest. I was slightly offended by your distrust towards the Russians, since I always play them. Scientific and expansionist is a good combo if, like me, you're not interested in a conquest or dominance victory, although on Deity and maybe Emperor you have to get a bit bloody at the early stages.

Your analyses are based on offensive tactics, which for me are not very interesting for the following reasons:

1) It seems to me like a cruder form of chess.
2) It takes longer for me to play, since I ponder a lot over what to do, and have a slow computer.
3) If I win - fine, but if I don't get anywhere with the war waging, it's frustrating like hell.
4) I prefer having relations with the other nations. Although I realize they are not people, I want tranquil looking folks to look at - just like a newborn.

Maybe I just don't see all the possibilites of the game - I've never used Communism for instance. On the other hand, isn't this a different game depending on which type of victory you want? It would be nice to hear your view on this.
 
I found your post interesting, but I disagree with a few of your analyses:

Persians: IMHO the easiest civ to play in the game (although I do agree, the Iroquois are pretty powerful). If you go for a despotic whipping game, then there's no contest. Even with some losses, a huge stack of Immortals are unstoppable in the ancient age. No retreat can be annoying, but with an attack of four, you still aren't going to take many losses.

Chinese: In my opinion, what makes cavalry such a great attacker (most of my games end in an industrial age attack with cavalry) is not so much the attack value, but the 3 movement which allows for a blitzkrieg attack. Before the AI's even know what's hit them, they've got cavalry ROLLING over their cities. The same applies to riders.

Germans: Although as I said above, most of my games are decided with cavalry, I still think panzers are a formidable opponent. I already explained above why a movement of 3 can be so deadly, and panzers combine that with the ability to attack more than once. If your opponents don't have rubber (or sometimes even if they do!) and you get panzers, watch them disappear within a few turns.

Good post though. :goodjob:
 
I think you're being way too hard on the Russians. Defense can be good for fast offensive units, because after they attack they will likely be attacked. In fact that's the reason you liked the Zulus, because they can guard fast offensive units after they attack.

You keep saying that 1-move units are bad because they can't retreat. What good is retreating? So you have a 1-hp, non-defensive unit in enemy territory(since of course this is an attacking unit), who will take a while to heal since it's in enemy territory, and is obviously close to a formidable enemy(because the enemy just kicked his ass!), and with retreating you lose the oppurtunity for a 'miracle comeback' in which at 1 hp you bring the other guy down and kill him(it seems to happen more often to the AI, but it has happened to me before).

English - sure, a world map only has value so long, but keeping them poor by selling it to them for exorbitant prices every few turns cripples them and loads up your pocket and/or your science early in the game, giving you an early lead which is very important.

I'll kind of agree with the Persians, but the Immortals are worthless on an archipelago map. I just started a game as the Persians, planning to expand early and storm with my Immortals in the ancient age. This is on the real world map. Where do I end up but......Japan! And the closest two civs to me were the Indians who started in Saudi Arabia/Iraq and the French who started in the US. By the time I met anyone I had knights. Damn game :)
 
Hail thee, Arathorn father of Aragorn!

I have a big problem with the Aztects: What is the use for Jaguar Warriors? Do you really bother to build them. I mean, what do you need a 1 attack unit for? Even in the stone age every opponent has at least spearmen. Ok, the retreat may work well sometimes. BUT: If the unit retreats and you can't move it away it's dead (1 defense?). No change. Also, do you ever attack with stone age units? I always wait until I have swordsmen or Horsemen (or War Chariots, which I think are soo good that they make the EGYPTIANS the easiest nation).
 
I disagree about the Aztecs. Even if you don't use their Jaguar Warriors for combat, they basically count as free scouts even though the Aztecs don't have the expansionist ability. Plus, the Aztecs don't have to worry about their Jags being ambushed by barbarians - the Jaguar Warriors can defend.

If you do use the Aztec Jaguar Warrior for combat, they can be formidable opponents, even against spears. If you go for really early combat, there are no swordsman on the board, and simply by building a lot of Jaguar Warriors, you can overwhelm even staunchly defended cities in the early ancient age. In one of my games (Monarch, large map, continents) I took out my neighbors the French within the first 10 to 15 turns with lots of Jaguar Warriors. Sure, you get an extremely early golden age, but that only serves to drastically up the rate at which you pump out your Jags. Plus, having a religious ability never hurts, and being militaristic ups the chances of your Jags spawning a GL with all that early combat, and a GL that early can be very decisive.
 
I should have stated this in my first post. I am posting looking at the following set-up:

Difficulty: Deity (maybe Emperor)
World: Standard size, everything else random

The very high difficulty level has forced me into reasonably set tactics to have a chance of winning. I've found that the civs high on my list work well. On easier levels, other civs might be easier.

The random world makes things interesting, too, because it makes expansionist a cr@p shoot, in that it might work very well, but you might be trapped on a small island and the scout is practically worthless.

Ohwell: You won the German SG because of superior tactics along the way. You would have won in an extremely similar fashion with ANY civilization. Despite being scientific, you were behind in science much of the game. The militaristic at the end surely helped, but once you figured out to target rubber/oil, the game was over.

Megalou: I play to win, however it comes about. In my deity games, I'm way behind in tech and it seems the only way to catch up is an ancient war. I *think* this is a feature of deity and not my play style, but I'm not certain. I'll happily win any way I can and usually don't decide until after the Industrial Age is over. Culture is, for me, the hardest winning condition, because I find larger maps very tedious.

Schnarrd: I've not played the Chinese much, so I don't know the benefit of moving three. My experience with MA wars, though, is that cultural borders aren't usually that large and if they are, you need to bring bombard units anyway. The ability to retreat to a barracks'd city might be critical. Just didn't strike me as so. And some civs have to be harder to play, else it's not a scale!

thcsquad: Retreat is awesome because the unit lives to fight another day. I rarely attack unless I expect to win. So that one unit might get his butt kicked and be sitting there, but it won't be enemy territory when my turn ends, unless I'm seriously SERIOUSLY screwed. And then it doesn't matter much what I have. (Retreat also keeps elites alive longer, increasing chances for Great Leaders.) As for the Russians, if I could upgrade knights to cossacks, it wouldn't be so bad, but losing that upgrade path relegates them to bottom-of-the-pile for me.

thamis: Just a note on Jaguar warriors...playing on diety, using pretty much exclusively the ultra-cheap jaguar warrior, I managed to defeat the Greeks (including a ton of hoplites), the Americans, and the Romans (in that order), gaining about 5 GLs in the process. I had a good 30-40 of those little jaguars running around, basically never dying. They are so cool because they are so cheap. So, yes, I attack with stone age units -- often because I think it's the best way to win the game.

Thanks for all the thoughts and comments!!!

Arathorn
 
I actually really like the Russians in Civ3.

IMO Cossacks are still a good units, even though they cannot be upgraded to or from. With 6 attack, 4 defense, they are basically musketeers and cavalry in one, which makes it very easy to hold captured cities, as you don't need extra 1MP defenders.

In the games I played, Cossacks were still useful upto Modern Armour, since Cossacks have 3 move and tanks have movements of only 2, which is of course true of normal cavalry, but enhanced defense stregth allows Cossack to act independently. On good terrain they can even successfully fend off cavalry.

Being scientific, Russians gain half-price libraries and univerisities, which not only good for science but for culture as well. Free tech advances at beginning of era are also welcome (3 free advances!).

I agree they I don't see any advantage of Russians being expansionist.
 
Having played some games at deity now, I am also convinced that you have to wage war in the beginning. Only when I made an early conquest did I get into contention in those games (lost anyway).

In the beginning I want to be scientific in order to go directly for iron working and secure iron. But that is less effective now that v1.16 has increased max turns to discover to 40. I' ve never tried conquering a civ without first having iron.

As for cultural victory, I agree it seems boring. Certainly more so than conqeust.
 
Great thread, Arathorn. Some comments and questions:

1. I've started playing on Emperor, having notched my first Monarch victory a few days ago, so I'm starting to understand why people at Deity emphasize an early conquista. But must one play the entire Deity game at war -- conquer the world by the Middle Ages? Or is it possible to conquer one civ early, then catch up and eventually overtake the AI in tech (for spaceship) or culture? Or at least make enough headway to build the UN?

2. For that matter, is the early conquest an absolute must in Deity? What if you start alone on a Standard map, and you're lucky enough to find yourself with 3-4 nice luxuries that can be traded for tech etc?

3. I also like your 'easiest' choices. The Iriquois Mounted Warrior is one of the best unique units in the game. I'm also a fan of the Samurai; it's a great all-around unit. In my one victory as Japan (at Regent level), I kept them hanging around all the way til the endgame; they made for good fast garrisons, to quelch resisters. But I'm not so sure about the Aztecs and the Jaguar Warrior. Yes, they're cheap, but how do you support a sizeable army of the things so early? My early Despotic civs are so low on cash that I have trouble funding a big army of anything.

4. Finally, a question about upgrades. Is it really that important to be able to upgrade? When I first started playing, I was an upgrade fanatic: I always tried to build Leonardo's, and I never finished the game with Stone Age units. But as I've ratcheted up the difficulty levels, I've almost never been able to build Leonardo's. And, just as important, I've found it crucial to have a large number of military units around to deter the AI.
As a result, I tend to retain my old Spearmen and even Swordsmen as fodder for city garrisons, cheap units to squash resisters, low-risk garrisons in a newly-conquered city that might flip. That last role is especially useful: I hate to lose a Mech Infantry to a city flip, but I kinda rejoice when I get 3 Swordsmen or Spearmen off the books. <g> Upgrading is pricy. Only in emergencies do I ever upgrade anything. Am I crazy?
 
Good post, though I haven't found much of a huge difference in any of the civs. I don't think the differences are quantify-able enough to amount to say one civ is "much" harder than another.

I think it's easier to say that some civs are better suited for different victory conditions. For example, it would be a little more challenging to win a space race & not be scientific or dominate the world & not be militaristic. I'll admit, I like some UUs better than others, but the civ traits are pretty even, IMHO (a little map dependent though).

I'm not too caught up in the upgrading myself. For example, if I'm russia, & know I can't upgrade my knights, I tend to put off a war until I can make the cossacks. That way, I won't have all kinds of knights sitting around. If I do have knights, I agree with Grotius. They're good fodder/riot police/etc.

I'm sure everyone can chime in with a successful game with your hardest civ or a tough game with the easiest civ, but I would challenge your claim to Russia as a difficult civ. Being expansionist & scientific is a great attribute for a democratic space race (as Russia! some irony there :) ).

Being Russia in my last game, I got a total of 9 free techs thoughout the game - 4 through goody huts, 3 free at new eras, and 2 with the Evolution wonder. I also had a valuable map and techs to sell.

Their UU is pretty good too. 1 extra defense isn't the greatest, but the calvary is useful for a long long time (best offense until tanks). And with the 1 extra defense, I'm not scared of leaving them on a hill by themselves in enemy territory, allowing me for very fast blitzes.

I do have 1 question. You mentioned that the Iroquois have a great early Golden age but the Greeks and others have a terrible early GA. Would you clarify this?
 
I always play on huge maps, with 60% water. I used to play with the french or the americans.
Playing on regent level on these maps, with an expansionist civ, will give you a huge tech lead, which the AI will never catch back. Basically, I always end up with an "Age" of advance.
I have never tried the russians, but I believe I could use the free techs ;)
Anyway, for me, industrious is a must have. Last time I played, I used the egyptians and the high speed temples gave me a cultural edge I never lost. But I ended up having a lot more trouble kicking butt because I wasn't as advanced as I used to be when playing with an expansionist civ.

loki
 
First, great post!

I find Persians easiest to play mostly because of my playing style.

I am very agressive early so I can steal cities and grab more territory than the AI with my immortals. They clearly rule the ancient era! For me, the early warfare is critical because I weaken my neighbors, and carve out a large contigous empire for later. Once that is accomplished, I have many paths to victory depending on the map, resources, etc. But my favorite victory is domination or conquest (I think I need to see someone about that) and this is a great start!

Since I do most of my fighting early, I have plenty of time to repair my relations with these nations later if I choose. But one thing is for sure, my neighbors will never again be a threat and will act as a buffer between me and stronger nations.

The other part that is great about Persians is the fast workers. I hate using non-industrious civ's, I want to build fast and waste as little resources as possible on workers that could better be used to build temples, barracks, and military units.

I'm not sure if Science helps all that much for my playing style. What would really be nice is Immortals + Religious + Industrious. But the fun of the game is the balance, if one civ has all the best traits, there's no decision.

My second favorite is the Romans just because it was the first I played and Legionare's are my second favorite UU. And they just seem like they should rule the ancient world!
 
Good post, could be part of a guide. I didn't concentrate that much on the upgrade paths until now, disbanding is another option.
My favorites are the Japanese. The Samurais really rule for a long period, enough to take the lead.

:goodjob:
 
My comments:

Grotius-
1: You don't have to play the entire game at war. It's all about catching up. Once you've achieved parity (or even close) in some of the important areas (production, land area, population, production, technology), LOTS of paths are open to you. It's all in that initial catch-up. As an example, I caught the world mid-Industrial Age in my current game and expect to win by space race.
2. I don't know. It has been for me, so far, but I still have lots to learn about playing the game better.
3. Jags are cheap. I ran at zero science and conquered a fair number of cities. In despotism, every city will support 4 units for free, so you can have a reasonably large army.
4. I am usually in Republic/Democracy by late-game and I hate supporting lots of relatively useless units. That said, I do keep all my living swordsmen around for garrison/resistance duty. I do pay for upgrades (except to the worthless musketmen), because I'm usually selling everything (luxuries, tech, etc.) I have acquired to get gold to upgrade/rush improvements. <Shrug> about which is better, though.

chiefpaco - I've never had that much luck with goodie huts and tech. Odds are you would have gotten 6 or so of those from any scientific civ, though (the three free age ones, the two from the wonder and about one goodie hut). Plus, the UU semi-forces you NOT to fight a war with knights, which is a detriment, instead of a strength, IMO. As for the GA thing, I was trying to present a general counter-point to the consensus opinion of "early GA bad", which shouldn't have been Iroquois specific. Mea culpa.

loki - Huge regeant maps may well have a different order or difficulty. I was writing specifically, if non-obviously, for deity on a standard map (I would GUESS it would be similar for small/large and might be greatly different for huge/tiny). Egyptians or Babylonians would probably be my recommendation for regeant and/or learner games.

There doesn't appear to be too much disagreement with most of the ideas, though. And just maybe somebody (else) learned something. (I know I did, just in making the list.)

Arathorn
 
I have to agree that the Persians should be rated higher. The Immortals are truly powerful attackers at the time when they are built. Granted, they can be slow and don't upgrade, but they really have helped me carve out a big chunk of land in the all-important early game. Not much can stand up to them. I understand your emphasis on mobility, but I feel that a 4 attack when even the best defender will only be a 3 is huge.

Also, Industrious has been the most useful civ trait for my games. It allows incredibly fast terrain improvements, so my cities don't have to build as many workers, allowing them to pump out settlers. Plus, it speeds up your tech by giving you roads on every square. Finally, it allows you to connect your empire with roads very quickly.
 
Originally posted by 7ontheline
I have to agree that the Persians should be rated higher. The Immortals are truly powerful attackers at the time when they are built. Granted, they can be slow and don't upgrade, but they really have helped me carve out a big chunk of land in the all-important early game. Not much can stand up to them. I understand your emphasis on mobility, but I feel that a 4 attack when even the best defender will only be a 3 is huge.
I tried Immortals and they are really too slow. Unfortunately my neighbors in my first Deity game were the Aztecs and they tought me the superiority of the Jaguar Warrior. Elite Jaguar Warriors are multiple purpose units(good Scouts!). The Persians aren't well balanced overall, either. To take advantage of their characteristics I have to develop Literacy and their Barracks are expensive compared with those of the Aztecs.

Once again my kudos to Arathorn.:)
 
Top Bottom