Recently, the online party in the Juvenis campaign miscalculated the depth of the dungeon they were entering and found themselves unexpectedly short of food ... and to deal with the problem as a DM, I found myself turning to a very old methodology that I'd developed more than two decades ago. All this time, I had never actually built a set of rules to manage starvation. As I write this, the trial has passed for the party, but I find myself wanting to settle on a hunger-starvation system nonetheless.
The principle problem is that it can take humans an extremely long time to die by starvation, particularly if they are able to get at least a little food. For example, 19 men survived a journey of 1,500 miles on an ice floe for six months at the close of the disastrous Polaris expedition, from 1871 to 1873. These are men who had been on bare rations for a year prior to the loss of their ship Polaris, who survived by straining just enough shrimp from the sea to give them each as little as an ounce a day, supported by the killing of a very rare seal, which might be gotten once in three to four weeks. And a seal does not go far between 19 men.
Some people do just give out ... but many will survive, though they may not find the energy to act. The goal, then, is to not to build a system of saving throws, in which the character simply dies, but a slow, grinding means of eventually killing a character off, while giving the potential for that character to survive a long, long time.
My idea is that the character would suffer acutely right at the start. That going without food initially has a severe effect, dropping the character's ability stats badly on the first day of feeling the pinch, and an equal amount on the second day. However, while this penalty doesn't go away, the severity of the increase diminishes as time passes.
Suppose the character is reduced to acting on half rations, and we want to know precisely what sort of effect that will have. I propose the table on the right as an initial proposal ~ there's more to come on the subject, but I want to be sure this part is understood first.
"Half-rations" is defined as an insufficient amount of food, below the amount that is required, but more than half that amount. If we suppose the character needs 2 kg of food in a given day in order to do heavy work, spread between two or three meals, then half rations would be anywhere between 1 kg and less than 2 kg. How much less is really up to us; but for game purposes, we don't want to reward the clever player who figures out that we will handwave two grams below the amount needed. We should then consider the numbers to be absolute.
Still, if we want to quibble, there's room for it on this table. If a day of half rations lowers your ability stats by 10%, we can argue that 1.9 kg should reduce the stats only 1/10th of 10%, or a mere 1%. It depends on how fine we want to split that hair. We can argue that even 1% less than a 17 is 16.83, which we can state is a defacto 16 for game purposes. That would discourage players from splitting hairs ... but we can go full game, too, and say that any deficiency in the amount of food is enough to cut stats the full 10%. I personally lean to this approach, as I just don't want to reward cheap players.
Okay, let me explain the table. As I say, that 1st day hurts, as does the 2nd day. Both see a severe drop in the character's stats. But then the idea of living with hunger begins to take hold, so that the 3rd and 4th days cause an additional penalty of only 5% each. The 5th, 6th and 7th days add a penalty of only 3.3% each. And finally the 8th through 12th days cause further penalties of only 2% per day.
I am extrapolating this on the Fibonacci series (endlessly useful), so that further 10% segments are successively divided by 8, 13, 21, 34 and 55. Altogether, that would create a starvation that would kill everyone in 143 days, or just under five months, somewhat crueler than the measure of the Polaris expedition survivors, but in line with other, similar experiences, such as Shackleton's adventure or John Franklin's first expedition ~ the one he wasn't lost in ~ overland through the Northwest Territories. However, this doesn't take into account the effects of eating less than half rations. How would that work?
My choice would be to count quarter rations up to half rations as two days of starvation; and then to count one-eighth rations up to a quarter rations as three days. Anything less than one-eighth would count as four days. That would bring starvation around a little faster than the Polaris expedition, but then we're not dealing with the healthier 19th century human. In any case, one could live a long, long time on one-eighth rations.
My food rules say that a sedentary character, one who needs to do little more than rest, make food for themselves and manage the small duties of living in a camp, must eat two pounds of food a day (my 17th century system uses imperial units). One-eighth of this is a mere 4 oz. per day. If the character tried to do hard work, that would require twice as much; and if the character were to take a part in battle, three times as much. But let's say we're sedentary characters, with little food, waiting to be rescued before we all die of starvation. Or by some other means.
How would this work, exactly. Well, that's a very long table ... but let's go as far as 33 days. From day 13 to day 20, the character is losing 1.25% of their stats each day, and from day 21 through day 33, the character is losing only 0.77% of their stats each day:
Here we can see the days of reduction to stats applied to a real character, that of Rob Munro the Scot from my Juvenis campaign, a druid. The numbers drop precipitously at the start, quickly flattening out at the high scores drop into the sixes and sevens after two weeks. The reader can see that I've highlighted the ability stats as they drop, in lighter orange for the most part, and under the dexterity at one point, in a darker tone.
The first shading indicates the point at which the druid no longer has the necessary stats to act as a druid. Being old school regarding D&D stats, the druid needs a 12 wisdom and a 15 charisma; and every other stat has to be 6 or higher, as the old Player's Handbook indicates. Here, the character's wisdom, dexterity and charisma all fall off the minimum on the third day. That is sobering.
On the 19th day, the character's dexterity falls below 3, typically viewed as the minimum roll for any character. I see that as significant, as the point in which the character must make an ability check to perform any dexterous activity, even walking or feeding themselves. And since that check is going to fail 90% of the time, this particular character is going to suffer very badly from any long-term starvation. That 7 dexterity is a harsh disability. Still, there's nothing that says a check to see if the character could walk can't be done as often as necessary, noting the time this would spend as the character eventually got the strength to rise, only to collapse again a few rounds later.
That brings me to the subject of ability checks in general. When the party was starving in the game of late, I had the active members all make an ability check against a random stat, once per day. For example, I had the druid here make a check against constitution, which the character succeeded. On some level, because the characters were following a route through a familiar, cleared dungeon, I could have asked for more than one check (as it is a dangerous place) ~ but normally, I wouldn't ask for any checks unless the players did something purposefully dangerous. My logic was that they did not "feel like themselves," and that this justified at least one check.
I think I'd increase the number of checks being made per day over time: perhaps 2 checks per day on the start of the fifth day, then three checks starting on the 9th day, and so on. After a while, though the characters could manage to stay alive, they'd yet become virtually helpess ... particularly as every stat dropped below a 3.
That 3 could be treated as another threshold, where constitution was concerned. Breathing is a constitutional action. Just as the heart beating is a strength issue; or being able to think is an intelligence requirement. At some point, having to be constantly making a roll for these things ought to end in the body just giving up the ghost.
The question is when, or how. I haven't quite worked that out yet. I like this scheme so far ... but it still has holes that I need to fill.