Because the SAT and ACT each have over one-hundred questions and students feel pressured to finish in time, it is very hard for almost all students not to make some silly mistakes. When I take the official tests (as I do every year), sometimes I make one or two mistakes as well. So, what is the margin for silly mistakes on both the SAT and ACT? And how does that difference in margin affect students’ testing? Let’s explore.
The SAT provides a significantly smaller margin for error (i.e. it penalizes any wrong answers – and thereby silly mistakes – much more 
heavily). How much more heavily? On average, each question wrong on the SAT brings a student’s score down 30% more than each question wrong on the ACT. Most of that difference is simply a function of the number of questions on each exam. On the SAT, there are 152 questions in total; on the ACT, 215. That’s 31% more questions on the ACT, which means that each question on the ACT is also a smaller portion of a student’s overall score, so students can miss more questions without as large of an effect on their scores. Here are some comparisons between SAT and ACT curves and their impact on students’ scores.
We analyzed the curves on the six most recently released SATs and ACTs and then compared them. (All standardized tests are curved so that students aren’t penalized by lower scores when they take a hard test and aren’t helped by higher scores if they take an easy test.)
Perfect Score Comparison
Verdict: The worst curved ACTs (up to 3 questions wrong to get a perfect score) still had a better curve than the one SAT with the best curve (up to 2 questions wrong to get a perfect).
Math Section Comparison
Verdict: Students could miss the same number of questions – 9 of them – on the average ACT Math section as they could on the best 
curved SAT Math section in order to get the same score. Additionally, because the ACT Math section only accounts for ¼ of the entire ACT composite score and the SAT Math sections account for ½ of the entire SAT score, we would really need to double the margin of error for the ACT in order to give that section equal weight and fairly equate it to an SAT score. Thus, in comparison, the ACT Math section has a dramatically better curve than the SAT Math section.
Verbal Section Comparison
Verdict: Again, the worst ACT curve (13 questions wrong for a 31) provided the same score as the best SAT curve (13 questions to score a 700).
Conclusion
This analysis provides more clarity as to why ACT scores are more consistent and predictable.
Because the SAT, on average, has a 30% steeper scoring curve than the ACT, students’ scores fluctuate significantly more from test to test. 
In essence, if every score is a step in a staircase, the SAT’s staircase has 30% fewer steps. So, if you get one more question wrong, you drop a 30% farther distance in score than on the ACT. Thus, if a student has a bad day and makes a few silly mistakes on the SAT, they could see their score drop 30-70 points (equivalent, at the top of the SAT’s scoring scale, to 1-2 full composite points on the ACT). In contrast, because the ACT has a better curve, a bad day with a few silly mistakes will likely only drop a student’s score by half a composite point or a full point at most.
Natural and expected variability between test scores is, as we wrote about here, important for students to understand so that they don’t get dejected by a lower than expected score. With less margin for error on the SAT than on the ACT (especially at the high end of the scoring scale), students taking the SAT need to prepare themselves for more variability in their scores.
