In an effort to solve a completely unrelated problem, I put a little something on an Excel sheet. It should be transcribed correctly here (unless it's not).

With

• $p$, the probability it backfires
• $n$, the number of uses
• $q$, the probability it backfires at least once within $n$ uses

It gives

• $q=1-(1-p)^n$ , as I said the probability it backfires at least once within $n$ uses
• $n=ln(1-q)/ln(1-p)$ , the number of uses at which there is a probability $q$ that it backfires at least once

Assuming $p=0.1\%$, after 100 uses, you have a 10% probability to end with at least one pudding monster. After 1000 uses, you're at 66%. After 3000 uses, you're at 95%.

Needless to say, you will need a pudding monster defence budget, if you are going to use this device with any sort of regularity.

For transporting goods, it would be fairly easy to just isolate the device by putting it in a big oven-like room/building, check arrivals for ravenous monsters and activate the defence system when necessary. This may damage the device, and your insurance premium will be quite high. But it might be cheaper than the cost of exploding rockets or refining unobtainium fuel for spaceships. Decisions, decisions.

For transporting humans, a typical Boeing 747-8 seats 467 people according to Wikipedia. So, assuming you transport people one by one (or that the probability to be affected is per person rather than per trip), you have a 37% chance that at least one of your passengers turns into a bloodthirsty creature from hell that may or may not kill the 466 other. You decide if you want to buy that ticket.

In an effort to solve a completely unrelated problem, I put a little something on an Excel sheet. It should be transcribed correctly here (unless it's not).

With

• $p$, the probability it backfires
• $n$, the number of uses
• $q$, the probability it backfires at least once within $n$ uses

It gives

• $q=1-(1-p)^n$ , as I said the probability it backfires at least once within $n$ uses
• $n=ln(1-q)/ln(1-p)$ , the number of uses at which there is a probability $q$ that it backfires at least once

Assuming $p=0.1\%$, after 100 uses, you have a 10% probability to end with at least one pudding monster. After 1000 uses, you're at 66%. After 3000 uses, you're at 95%.

Needless to say, you will need a pudding monster defence budget, if you are going to use this device with any sort of regularity.

For transporting goods, it would be fairly easy to just isolate the device by putting it in a big oven-like room/building, check arrivals for ravenous monsters and activate the defence system when necessary. This may damage the device, and your insurance premium will be quite high. But it might be cheaper than the cost of exploding rockets or refining unobtainium fuel for spaceships. Decisions, decisions.

For transporting humans, a typical Boeing 747-8 seats 467 people according to Wikipedia. So, assuming you transport people one by one (or that the probability to be affected is per person rather than per trip), you have a 37% chance that at least one of your passengers turns into a bloodthirsty creature from hell that may or may not kill the 466 other. You decide if you want to buy that ticket.

For comparison, this study on "worldwide fatal accidents to jet and turboprop aeroplanes above 5,700kg engaged in passenger, cargo and ferry/positioning flights for the ten-year period 2002 to 2011" (essentially big planes), page 30, gives a fatal accident rate of 0.6 per billion flight, or $0.00006\%$ which can be compared with the device's probability of failing. Taking $p = 0.1\%$ again, that's a difference of 4 orders of magnitude.

I chose air travel because it's probably the closest analogue to taking a spaceship to the Moon we have. Also I couldn't find data on fatal accident rate per trip for road transport.

In an effort to solve a completely unrelated problem, I put a little something on an Excel sheet. It should be transcribed correctly here (unless it's not).

With

• $p$, the probability it backfires
• $n$, the number of uses
• $q$, the probability it backfires at least once within $n$ uses

It gives

• $q=1-(1-p)^n$ , as I said the probability it backfires at least once within $n$ uses
• $n=ln(1-q)/ln(1-p)$ , the number of uses at which there is a probability $q$ that it backfires at least once

Assuming $p=0.1\%$, after 100 uses, you have a 10% probability to end with at least one pudding monster. After 1000 uses, you're at 66%. After 3000 uses, you're at 95%.

Needless to say, you will need a pudding monster defence budget, if you are going to use this device with any sort of regularity.

For transporting goods, it would be fairly easy to just isolate the device by putting it in a big oven-like room/building, check arrivals for ravenous monsters and activate the defence system when necessary. This may damage the device, and your insurance premium will be quite high. But it might be cheaper than the cost of exploding rockets or refining unobtainium fuel for spaceships. Decisions, decisions.

For transporting humans, a typical Boeing 747-8 seats 467 people according to Wikipedia. So, assuming you transport people one by one (or that the probability to be affected is per person rather than per trip), you have a 37% chance that one of your passengers turns into a bloodthirsty creature from hell that may or may not kill the 466 other. You decide if you want to buy that ticket.

In an effort to solve a completely unrelated problem, I put a little something on an Excel sheet. It should be transcribed correctly here (unless it's not).

With

• $p$, the probability it backfires
• $n$, the number of uses
• $q$, the probability it backfires at least once within $n$ uses

It gives

• $q=1-(1-p)^n$ , as I said the probability it backfires at least once within $n$ uses
• $n=ln(1-q)/ln(1-p)$ , the number of uses at which there is a probability $q$ that it backfires at least once

Assuming $p=0.1\%$, after 100 uses, you have a 10% probability to end with at least one pudding monster. After 1000 uses, you're at 66%. After 3000 uses, you're at 95%.

Needless to say, you will need a pudding monster defence budget, if you are going to use this device with any sort of regularity.

For transporting goods, it would be fairly easy to just isolate the device by putting it in a big oven-like room/building, check arrivals for ravenous monsters and activate the defence system when necessary. This may damage the device, and your insurance premium will be quite high. But it might be cheaper than the cost of exploding rockets or refining unobtainium fuel for spaceships. Decisions, decisions.

For transporting humans, a typical Boeing 747-8 seats 467 people according to Wikipedia. So, assuming you transport people one by one (or that the probability to be affected is per person rather than per trip), you have a 37% chance that at least one of your passengers turns into a bloodthirsty creature from hell that may or may not kill the 466 other. You decide if you want to buy that ticket.